WO2009075554A2

WO2009075554A2 - Patent information providing method and system

Info

Publication number: WO2009075554A2
Application number: PCT/KR2008/007398
Authority: WO
Inventors: Min Soo Kang
Original assignee: Kwanggaeto Co Ltd
Current assignee: Kwanggaeto Co Ltd
Priority date: 2007-12-12
Filing date: 2008-12-12
Publication date: 2009-06-18
Anticipated expiration: 2010-06-12
Also published as: KR20090063156A; WO2009075554A3; KR101534155B1

Abstract

Disclosed is a method of providing patent information, which includes multi-dimensional patent analysis information. By the method, it is possible to obtain various types of patent analysis results.

Description

PATENT INFORMATION PROVIDING METHOD AND SYSTEM

[Technical Field]

The present invention relates to a method for providing patent information, and more particularly to a method and a system for providing patent information, which includes multi-dimensional patent analysis information.

[Background Art:]

Patent information has characteristics of technological information, rights information, and management information, and has an increasing importance in the globalized international competition. As technological information, the patent information enables the understanding of technology development trends and technical ideas applied to individual patents. As rights information, the patent information enables the understanding of the scope of each patent and the degree of domestic or foreign patent acquisition. As management information, the patent information enables the understanding of technology development trends of competing companies.

Patent information can be obtained through various methods, representative of which include 1) a method using a patent information system provided by a patent office of each nation, 2) a method using the patent information system developed by a non-governmental company, and 3) a method using a patent information system developed by an individual company based on the object of the company. Representative examples of the method of 1) include patent information systems provided by the Korean Intellectual Property Office (KIPO; www.kipo.go.kr) and the Korean Intellectual Property Rights Information Service (KIPRIS; www.kipris.or.kr), and representative examples of method 2) include www.delphion.com of the United States, www.patolis.co.jp of Japan, www.wips.co.kr and www.wisdomain.com of the Republic of Korea.

Non-governmental companies of each nation, such as Thomson Scientific operating www.delphion.com, possess their own constructed patent information database, and provide search results in various ways in response to a search formula input through a patent search engine. Further, there are various types of analysis software developed based on patent information as described above, among which especially the software, which has been developed by the company (which has been merged by Thomson Scientific) operating the site of www.micropatent.com and is spread with a trademark named j ° AlEEKA. i ± Further, most major worldwide technology- advanced companies have constructed and are operating their own internal patent search/management systems. However, it is difficult for an external user to access the internal patent search/management systems of those companies. Usually, the construction of such a patent search/managementsystem requires a huge expense more than thousands of millions of won, which prevents most domestic or foreign middle-sized and small-sized venture companies from constructing their own internal patent search/management systems. Many of these middle-sized and small-sized venture companies have exclusive patent management systems, which manage their own patents and collect patent information for the companies. However, many small- sized companies or companies not having a large concern often do not have such an exclusive patent management system. These companies usually entrust management of their own patents to patent service companies or law firms.

The exclusive patent management systems use on-line patent information service companies, such as www.delphion.com, which are accessible from outside, and collect and manage patent information through platforms provided by on-line patent information service companies. Further, it is a worldwide trend that inventions of most patent applications filed in the name of a company are made by inventors belonging to the company and those inventors collect domestic or foreign patent informationin relation to their own inventions. This is because more than 70% of worldwide new technology information is disclosed in the form of patent information in spite of existence of the patent publication term and patent information contains ideas more commercially available than other information including thesis information. However, many of these inventors are not experienced in patent search. Moreover, it is convenient for even inventors experienced in patent search to access external patent informationservice providing companies, such as www.delphion.com, and obtain patent information from them.

It is usual that inventors continuouslyengage themselves in research in a particular technological field for several years or several scores, and it is thus usual that there is not a big change in the field of inventions continuously made by an inventor. Therefore, domestic or foreign patent information of a field directly or indirectly relating to a technical field of an invention of an inventor would be very usuful to the inventor. For this reason, there has been a request for a distinguished patent information service system, which can be operated not only company by company but also inventor by inventor belonging to the individual company, so that it can provide an optimized service to each inventor. It is expected that such a patent information service system will largely improve the accessibility of patent information by middle-sized or small- sized venture companies, which are incapable of developing and maintaining the system by themselves, and will serve as a new platform for in-company circulation of patent informationfor the companies capable of developing and maintaining the system by themselves.

Further, if patents managed by companies or inventors are automatically analyzed, and analysis information on competing companies or competing technologies in relation to the patents and monitor information on the competing companies or competing technologies are automatically provided to the companies or inventors, it will be very convenient and useful for the companies or inventors.

[Disclosure] [Technical Problem]

A first object of the present invention is to provide various methods of pre-processing patent document information, a method of generating data or information necessary for patent analysis based on the pre-processed patent information, a method for various patent analyses through the generated information, and a patent information system thereof. A second object of the present invention is to provide a method for automatically constructing company-based, applicant-based, inventor- based, agent-based, and patent classification code-based patent information systems, and the systems thereof. A third object of the present invention is to provide a method and a system for information processing, which can provide monitor, search, directory, and analysis services automatically including lower patent classification codes of a to-be-processed patent classification code.

A fourth object of the present invention is to provide a method and a patent information system for providing patent information obtained by achieving the first to third objects to multiple users on line. [Technical Solution]

In accordance with an aspect of the present invention, there is provided a method of providing patent information by a system providing patent information to a computer of a user, the method including the steps of: obtaining selection information for a patent document set; specifying a to-be-processed patent document set based on the selection information; and generating processed patent information by performing at least one preset processing for the specified to-be-processed patent document set, wherein at least one patent classification code is used in at least one of the step of specifying the to-be-processed patent document set and the step of generating the processed patent information by performing said at least one preset processing, use of the patent classification code corresponds to processing of information including patent information of a first patent classification code and lower patent classification codes of the first patent classification code when the first patent classification code is given in at least one of the step of obtaining the selection information and the step of generating the processed patent information and corresponds to a preset level in a patent classification code system to which the first patent classification code belongs, and the patent classification code belongs to at least one of International Patent Classification (IPC) system, United States Patent Classification (USPC) system, File Index (FI) system of Japan, F-term (FT) system of Japan, and European Classification (ECLA) system. It is preferred that the preset level is a dot level patent classification code, and the dot level patent classification code includes a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code in the patent classification code system. It is also preferred that the dot level patent classification code includes a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subgroup or below when the dot level patent classification code belongs to the IPC system, includes a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subclass or below when the dot level patent classification code belongs to the USPC system, includes a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subgroup or below when the dot level patent classification code belongs to the FI system of Japan, includes a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes under a theme code when the dot level patent classification code belongs to the F- term(FT) system of Japan, and includes a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subgroup or below when the dot level patent classification code belongs to the ECLA system.

It is also preferred that, in the processing of information including patent information of the first patent classification code and lower patent classification codes of the first patent classification code, when the first patent classification code is a dot level patent classification code, at least one patent document in a patent document set including all lower patent classification codes of the dot level first patent classification code is subjected to information processing.

It is also preferred that the selection information is obtained by at least one of: 1) an obtainment method of receiving input of the selection information through a search formula from the computer of the user; 2) an obtainment method of receiving selection information, which is selected from bibliographical information included in the preset patent document presented in the computer of the user and is transmitted from the computer of the user; 3) an obtainment method of receiving selection information on a particular patent document set selected in at least one patent document set generated in advance by the user; and 4) an obtainment method of receiving at least one preset condition from the system.

It is also preferred that, in the step of specifying the to-be- processed patent document set, when one of the obtainment methods 1) to 4) has been executed, at least one first patent classification code is included in the selection information during execution of the methods 1) to 4), and the first patent classification code is a dot level patent classification code, a patent document set including all lower patent classification codes of the dot level first patent classification code is specified as the to-be-processed patent document set.

It is also preferred that a process of specifying the patent document set including all lower patent classification codes of the dot level first patent classification code as the to-be-processed patent document set includes the steps of: obtaining all lower patent classification codes of the dot level first patent classification code; and specifying the to-be-processed patent document set by using a search formula or query formula including the first patent classification code and all the lower patent classification codes thereof.

It is also preferred that individual patents of the patent document set include at least one patent granted a weight according to at least one preset weight granting policy, and the processed patent information is generated while reflecting the weight granted to the individual patent.

It is also preferred that the weight granting policy includes calculation of at least one of 1) preset counting information and 2) preset calculation value information in relation to the individual patent, a counting value for the preset counting information includes at least one of 1) a number of families in at least one of a first nation and a second nation, 2) a number of claims, and 3) a cited-by number, and the preset calculation value information includes at least one of 1) at least one evaluation score for an inventor of the patent, 2) at least one evaluation score for an applicant of the patent, 3) at least one preset calculation value for a patent classification code included in the patent of the inventor, 4) at least one preset calculation value for a patent classification code included in the patent of the applicant, 5) at least one preset calculation value reflecting financial information of the applicant, and 6) at least one preset calculation value for the patent classification code itself included in the patent.

It is also preferred that processed patent information corresponding to the patent classification code information is provided together when the processed patent information is provided to the computer of the user, and the processed patent information provides processed patent information corresponding to at least one direct lower patent classification code of the patent classification code in a patent classification code system including the patent classification codewhen there is a request for the processed patent information corresponding to at least one direct lower patent classification code of the patent classification code from the computer of the user.

In accordance with another aspect of the present invention, there is provided a method of providing patent information by a system providing patent information to a computer of a user, the method including the steps of: obtaining selection information for a patent document set; specifying a to-be-processed patent document set based on the selection information; and generating processed patent information by performing at least one preset processing for the specified to-be-processed patent document set, wherein the selection information is obtained by at least one of: 1) an obtainment method of receiving input of the selection information through a search formula from the computer of the user; 2) an obtainment method of receiving selection information, which is selected from bibliographical information included in the preset patent document presented in the computer of the user and is transmitted from the computer of the user; 3) an obtainment method of receiving selection information on a particular patent document set selected from at least one patent document set generated in advance by the user; and 4) an obtainment method of receiving at least one preset condition from the system, the processed patent information includes at least one piece of mult i-dimensional analysis information generated through at least one preset analysis of the specified to-be-processed patent document set, dimensions of the multi^¬ dimensional analysis information include at least one of 1) time, 2) patent classification code, 3) applicant, 4) inventor, 5) agent, 6) nation, 7) keyword, and 8) convergence pattern, the multi-dimensional analysis information includes at least one preset counting value for at least one of the dimensions, a calculation value obtained by at least one preset calculation formula, a query result value obtained by at least one query formula, and a result value of processing by at least one analysis processing algorithm, for the to-be-processed patent document set.

It is preferred that, for patent classification codes included in the to-be-processed patent document set, the analysis information is generated by applying at least one of 1) an analysis method for analysis of all patent classification codes and 2) an analysis method for analysis of only main patent classification codes, and the analysis method for analysis of all patent classification codes generates analysis information by applying at least one of a) an analysis method of granting a uniform weight to all patent classification codes, and b) an analysis method of granting different weights to main patent classification codes and sub- patent classification codes.

It is also preferred that the analysis is an analysis of convergence between patent classification codes, which includes at least one of association analysis, frequent occurrence pattern analysis, correlation analysis, and grouping between at least two patent classification codes included in patent documents including at least two patent classification codes of an identical patent classification system from among the patent documents included in the to-be-processed patent document set. It is also preferred that, in the analysis of convergence between patent classification codes, when said at least two patent classification codes of the identical patent classification system are included in a single patent document and when said at least two patent classification codes are not highest patent classification codes or patent classification codes of a preset level, the analysis of convergence between patent classification codes is performed for each of said at least two patent classification codes together with higher patent classification codes thereof up to a preset level of each of said at least two patent classification codes. It is also preferred that the analysis of convergence includes at least one of 1) a convergence analysis between at least two technical keywords generated based on text information extracted from at least one preset field of one patent document, 2) a convergence analysis between co- inventors of a patent document having co-inventors, 3) a convergence analysis between co-applicants of a patent document having co-applicants, and 4) a convergence analysis between citation patent documents included in citation information of a single patent document, for individual patent documents of the to-be-processed patent document set, and the convergence analysis includes at least one of association analysis, frequent occurrence pattern analysis, correlation analysis, and grouping between at least two keywords, co-inventors, co-applicants, and co-citation patent documents included in the patent documents.

It is also preferred that the method further includes the step of visualizing a result of the convergence analysis through a network diagram including at least two nodes, wherein the visualized result include at least one of information on the nodes and information on lines interconnecting the nodes.

It is also preferred that the information on the lines interconnecting the nodes includes numerical information, and the method further includes the steps of-" obtaining selection information on the numerical information or the lines from the computer of the user; extracting root patent document information corresponding to the numerical information from a patent information database including the to-be- processed patent document set; and transmitting the extracted root patent document information to the computer of the user.

It is also preferred that the method further includes the steps of: obtaining change information of the to-be-processed patent document set from the computer of the user; and specifying the to-be-processed patent document set based on the obtained change information, wherein the obtaining of the change information includes'• providing an interface for the obtaining of the change information to the computer of the user; and obtaining selection information of the user for the interface, the interface includes at least one of 1) time type selection interface, 2) time interval selection interface, 3) nation selection interface 4) patent classification code selection interface 5) citation specification interface, 6) family information-including selection interface, and 7) after-search selection interface, the time type selection interface includes a selection interface providing selection of at least one of application date, publication date, and registration date, the time interval selection interface includes a selection interface providing selection of at least one of year/month/week/day of each of a starting period and an ending period, the nation selection interface includes a selection interface providing selection of at least one between at least two nations, the patent classification code selection interface includes a selection interface providing selection of at least one between a patent classification code type selection interface, a patent classification code level selection interface, and a main patent classification code and sub- patent classification code selection interface, the citation specification interface includes a selection interface providing selection of at least one of a citation direction interface for selection between a backward citation and a forward citation, a citation depth interface for selection of a citation depth, a citation selection interface for selection between a direct citation and an indirect citation, the family information- including selection interface includes a selection interface for at least one of determination if family information is included and selection of a nation range, and the after-search selection interface includes a processing interface for presenting a preset field for search in order to find the to-be-processed patent document set, receiving a search phrase through the preset field from the computer of the user, providing a search result corresponding to the search phrase, receiving at least one selection information selected in the search result, and determining only a patent document set relating to the selection information as the to-be- processed patent document set .

It is also preferred that the analysis is a set-based citation analysis that treats at least two patent documents as one analysis unit, the set-based citation analysis includes at least one of 1) backward citation patent document set analysis, 2) forward citation patent document set analysis, and 3) citation occurrence self patent document set analysis, the analysis of 1) to 3) creates at least one preset multi-dimensional analysis information from the patent document set, which is an object of the analysis of 1) to 3), a backward citation patent document set specified for the backward citation patent document set analysis includes a document set including citation information of at least one citation depth included in the to-be-processed patent document set, a forward citation patent document set specified for the forward citation patent document set analysis includes a later-filed document set including citation information of at least one citation depth citing the to-be- processed patent document set, a citation occurrence self patent document set specified for the citation occurrence self patent document set analysis includes a patent document set including patent documents cited at least one time from among the patent documents of the to-be-processed patent document set.

It is also preferred that, in the set-based citation analysis, specification of the backward citation patent document set includes specification of the patent document set and an indirect backward citation document set of the patent document set and specification of the forward citation patent document set includes specification of the patent document set and an indirect forward citation document set of the patent document set, the indirect backward citation document set includes patent documents, which have application dates prior to application dates of to-be-processed patent documents of the to-be-processed patent document set and commonly cite patent documents cited by the to-be-processed patent documents, and the indirect forward citation document set includes patent documents, which have application dates later than application dates of to-be- processed patent documents of the to-be-processed patent document set and commonly cite patent documents cited by the to-be-processed patent documents.

It is also preferred that the multi-dimensional analysis information is transmitted in a form of a table including numerical information to the computer of the user, and the method further includes the steps of: receiving selection information on the numerical information from the computer of the user; extracting root patent document information corresponding to the numerical information from a patent information database including the to-be-processed patent document set; and transmitting the extracted root patent document information to the computer of the user.

It is also preferred that at least one of the step of specifying the to-be-processed patent document set and the step of generating the processed patent information by performing at least one preset processing uses at least one patent classification code, use of the patent classification code includes information processing of a first patent classification code together with lower patent classification codes of the first patent classification code when the first patent classification code is given in at least one of the step of receiving the selection information and the step of generating the processed patent information and corresponds to a preset level in a patent classification code system to which the first patent classification code belongs, and the patent classification code system includes at least one of the IPC, USPC, FI and FT of Japan, and ECLA. It is also preferred that the obtainment method of receiving selection information, which is selected from bibliographical information included in the preset patent document presented in the computer of the user and is transmitted from the computer of the user, includes the steps of: providing at least one of applicant information including at least one applicant, patent classification code information including at least one patent classification code, inventor information including at least one inventor, and agent information including at least one agent to the computer of the user; obtaining selection information on at least one of the applicant information, the patent classification code information, the inventor information, and the agent information from the computer of the user; and generating a preset multi-dimensional analysis result by processing a patent document set specified by the selection information as a to-be-processed patent document set, and providing the generated preset multi-dimensional analysis result to the computer of the user. It is also preferred that the system generates at least one medium multi-dimensional analysis result table for easily and rapidly creating a multi-dimensional analysis result according to the applicants, the patent classification codes, the inventors, or the agents, and the method further includes the steps of: providing at least one of at least one preset applicant-based multi-dimensional analysis list or menu, at least one preset patent classification code-based multi-dimensional analysis list or menu, at least one preset inventor-based multi-dimensional analysis list or menu, and at least one preset agent-based multi-dimensional analysis list or menu to the computer of the user; obtaining selection information on the multi-dimensional analysis list or menu from the computer of the user; and when an analysis list of the obtained selection information is obtainable from the medium multi-dimensional analysis result table, generating an analysis result of the obtained selection information by referring to the medium multi-dimensional analysis result table and providing the generated analysis result to the computer of the user.

*37

*In accordance with another aspect of the present invention, there is provided a method of providing patent information by a system providing patent information to a computer of a user, the method including the steps of: obtaining patent document information on at least one patent document published in a first nation; when more than one year has passed after an application date of the published patent document in the first nation and a date on which the patent document information is obtained, referring to at least one of 1) PCT application information of the published patent document, 2) second nation family information claiming a priority to a second nation based on the published patent document of the first nation, and 3) current status information of a second nation patent document corresponding to the second nation family information; determining if the published patent document corresponds to a first case in which the published patent document has no PCT application information and no second nation family information, a second case in which the published patent document has PCT application information but has no second nation family information, and the application date of the published patent document in the first nation corresponds to a date after passage of a preset term from the date on which the patent document information is obtained, or a third case in which the published patent document has PCT application information and second nation family information, and the current status information of the second nation patent document corresponds to at least one of cancellation, expiration, final rejection, and final invalidation; and when the published patent document corresponds to one of the first case, the second case, and the third case, transmitting patent document information of the first nation and preset display information to the computer of the user, wherein the preset display information indicates that the published patent document corresponds to one of the first case, the second case, and the third case and the preset display information is displayed around the patent document information of the first nation when the patent document information of the first nation is displayed, wherein the published patent document has at least one second nation. [Advantageous Effects]

The present invention has the following effects. First, through various pre-processings, it is possible to efficiently construct a patent information database from patent-related root data, such as patent document information or patent classification code data.

Second, only by obtained applicant, inventor, and agent information, it is possible to generate a hierarchically incorporated applicant name-based patent information system, inventor name-based patent information system, and agent name-based patent information system. Third, through pre-processing patent classification code information, it is possible to efficiently process lower patent classification codes in patent search, various patent analyses, patent monitor, etc., which improves the quality of the patent search, patent analysis, and patent monitor service. Especially, by generating patent information automatically including lower patent classification codes, it is possible to improve the quality of the generated patent information.

Fourth, by processing convergence information on patent classification codes and technology keyword information, it is possible to understand the trend of convergence between patent classification codes and technology keywords and to use the convergence property in the patent analysis of the present invention.

Fifth, by performing various analyses including a citation analysis for each patent document set, it is possible to process patent information regardless of the size, type, etc. of the document set. Sixth, by analyzing co-applicants, co-inventors, accompanied prior-filed citation patent, accompanied patent classification codes, and accompanied technology keywords, it is possible to obtain a network relation between applicants, inventors, etc., a network between citation patents, a network between patent classification code, and a network between technology keywords, and to generate an analysis result for each of them.

Seventh, it is possible to determine if it is possible to exercise the exclusive right of a patent document, which is found as a result of search in a first nation, in a second nation, and to provide a basis for determination about the availability of the technology information included in the patent document of the first nation in the second nation. [Description of Drawings]

The foregoing and other objects, features and advantages of the present invention will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings in which:

FIG. 1 illustrates a construction of a patent information system according to an embodiment of the present invention;

FIG. 2 illustrates a DB unit according to an embodiment of the present invention;

FIG. 3 illustrates a patent document master DB according to an embodiment of the present invention;

FIG. 4 illustrates a patent classification code master DB according to an embodiment of the present invention; FIG. 5 illustrates a subject master DB according to an embodiment of the present invention;

FIG. 6 illustrates a multi-dimensional analysis operation result table DB according to an embodiment of the present invention;

FIG. 7 illustrates a support DB unit and a secondary processing DB unit according to an embodiment of the present invention;

FIG. 8 illustrates a pre-processing module according to an embodiment of the present invention;

FIG. 9 illustrates a master DB generation module according to an embodiment of the present invention; FIG. 10 illustrates a citation information pre-processing module according to an embodiment of the present invention;

FIG. 11 illustrates a patent information processing basic module according to an embodiment of the present invention;

FIG. 12 illustrates a support module according to an embodiment of the present invention;

FIG. 14 illustrates a patent intelligence module according to an embodiment of the present invention;

FIG. 15 illustrates a directory generating module according to an embodiment of the present invention! FIG. 16 illustrates an operation of a patent classification code search module according to an embodiment of the present invention;

FIG. 16 illustrates an analysis result for data of the number of applications of each year for each numerously-filed IPC of an IPC sub^¬ class level filed in the Republic of Korea by Samsung Electronics Inc.; FIG. 18 illustrates an analysis result generated by a drill-down to HOlL according to an embodiment of the present invention;

FIG. 19 illustrates an analysis result generated by a drill-down to HOlL 21/00 and lower patent codes thereof according to an embodiment of the present invention; FIG. 20 illustrates an analysis result for numerously-filed IPC- based each-year applications-number data of the IPC main group level by Samsung Electronics Inc. from among all applicants within a DB kept by a patent information system according to an embodiment of the present invention; FIG. 21 illustrates an analysis result for numerously-filed IPC- based each-year applications-number data of the IPC 1 dot sub-group level by Samsung Electronics Inc. from among all applicants within a DB kept by a patent information system according to an embodiment of the present invention; FIG. 22 illustrates an analysis result for numerously-filed IPC- based each-year applications-number data of the IPC sub-class level based on application documents filed in the US by Samsung Electronics Inc. from among all applicants within a DB kept by a patent information system according to an embodiment of the present invention; FIG. 23 illustrates an analysis result for numerously-filed IPC- based each-year applications-number data of the IPC sub-class level based on registered documents registered in the US by Samsung Electronics Inc. from among all applicants within a DB kept by a patent information system according to an embodiment of the present invention; FIG. 24 illustrates an analysis result for numerously-filed USPC- based each-year applications-number data of the USPC no dot (sub-class) level filed in the US by General Motors from among all applicants within a DB kept by a patent information system according to an embodiment of the present invention; FIG. 25 illustrates an analysis result for numerously-filed USPC- based each-year applications-number data of the USPC 1 dot level filed in the US by General Motors from among all applicants within a DB kept by a patent information system according to an embodiment of the present invention; FIG. 26 illustrates analysis and drill-down of total applications in relation to IPC H04B based on application documents filed in the Republic of Korea according to an embodiment of the present invention;

FIG. 27 illustrates analysis and drill-down of total applications in relation to IPC H04B based on application documents filed in the US according to an embodiment of the present invention;

FIG. 28 illustrates analysis of total applications of each numerously-filed company in relation to IPC H04B based on application documents filed in the Republic of Korea according to an embodiment of the present invention; FIG. 29 illustrates analysis of occupancy-based numerously-filed applicants in relation to IPC H04B based on application documents filed in the Republic of Korea according to an embodiment of the present invention; FIG. 30 illustrates analysis of activity index-based numerously- filed applicants in relation to IPC H04B based on application documents filed in the Republic of Korea according to an embodiment of the present invention;

FIG. 31 illustrates total application amount analysis including a drill-down of IPC H04B and lower classifications thereof based on US application documents according to an embodiment of the present invention; FIG. 32 illustrates a representative competing company analysis based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system of the present invention according to the exemplary embodiment of the present invention; FIG. 33 illustrates a representative competing applicant analysis for each numerously-filed patent classificatio code based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system of the present invention according to the exemplary embodiment of the present invention; FIG. 34 illustrates a representative competing applicant analysis based on an IPC main group of the total amount of the Samsung Electronics among all the applicants in US in the DB of the patent information system of the present invention according to the exemplary embodiment of the present invention; FIG. 35 illustrates a numerously-filed inventor-based analysis result according to the year based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system 1 of the present invention according to the exemplary embodiment of the present invention; FIG. 36 illustrates a numerously-filed inventor-based analysis result for each year based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system of the present invention according to the exemplary embodiment of the present invention; FIG. 37 illustrates a USPC sub-class (no dot, just lower than class)-based numerously-filed applicant ranking analysis result based on the total amount of US applications of the Samsung Electronics among all the applicants in US in the DB of the patent information system of the present invention according to the exemplary embodiment of the present invention;

FIG. 38 illustrates a year-based analysis result of citation total amount for the to-be-analyzed document set according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set;

FIG. 39 illustrates a year-based analysis result for the multiple- citing applicant according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set; FIG. 40 illustrates a year-based analysis result for the multi- citing patent classification code (IPC main group level) according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set ;

FIG. 41 illustrates a year-based analysis result reflecting the drill down of the multi-citing patent classification code (IPC main group level) according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set;

FIG. 42 illustrates a year-based analysis result of the multi- citing inventor according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the

Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set;

FIG. 43 illustrates a year-based analysis result of the most citing applicant with respect to the to-be-analyzed document set according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set; FIG. 44 illustrates a year-based analysis result for the most citing inventor according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set;

FIG. 45 illustrates a year-based analysis result when the IPC in the analysis for the most citing technology according to the IPC main group is drilled down according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set;

FIG. 46 illustrates a year-based analysis result when the IPC in the analysis for the most citing cited technology of the USPC sub class (no dot, class direct subordinate) is drilled down according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set;

FIG. 47 illustrates a total amount reference analysis result and a chart of the analysis result generated by a chart generating module among a reporting module of the present invention for the total amount reference analysis result with respect to the to-be-analyzed document set in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system are considered as the reference document set, the entire forward citation document set serves as the to-be-analyzed document set; FIG. 48 illustrates the analysis result of a year-based citation total amount of the often-cited inventor and a document list generated by a simplified analysis module that provides the simplified analysis result (document list, the number of applications/registrations of the most- filing applicant according to the year, the year-based number of applications/registrations according to the most-filing inventor, the number of applications/registrations of the most-filing technical field (IPC, USPC, FT) according to the year (including drill-down) for the document set corresponding to the specific number when the specific number represented in the analysis result is clicked, with respect to the to-be- analyzed document set in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set; FIG. 49 illustrates that the simplified analysis module according to the present invention also provides a function of the drill-down according to the most-filing technology field (IPC, USPC, FT);

FIG. 50 illustrates a year-based analysis result of the citation total amount with respect to the to-be-analyzed document set, in which, when the numerously-filed IPC sub class unit document in all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set ; FIG. 51 illustrates a hierarchical patent informatoin service system, which is a sub-system of a patent informatoin system according to an embodiment of the present invention;

FIG. 52 illustrates an individual unit patent information system generating enginean according to an embodiment of the present invention; FIG. 53 illustrates an internal construction of a patent information pre-processing module according to an embodiment of the present invention;

FIG. 54 illustrates a weight pre-processing module according to an embodiment of the present invention; FIG. 55 illustrates a citation information pre-processing module according to an embodiment of the present invention;

FIG. 56 illustrates a patent classification code pre-processing module 301-3-1 or 3500 according to an embodiment of the present invention; FIG. 57 illustrates a representative applicant name pre-processing module according to an embodiment of the present invention!

FIG. 58 illustrates a representing phrase extracting pre^¬ processing module according to an embodiment of the present invention;

FIG. 59 illustrates a family information pre-processing module according to an embodiment of the present invention,"

FIG. 60 illustrates a multiple patent classification code correlation pre-processing module according to an embodiment of the present invention;

FIG. 61 illustrates a structure of the patent classification code- based statistic pre-processing module according to an embodiment of the present invention;

FIG. 62 illustrates a patent information intelligence module according to an embodiment of the present invention;

FIG. 63 illustrates an analysis module according to an embodiment of the present invention,"

FIG. 64 illustrates a patent information system arrangement generation engine according to an embodiment of the present invention;

FIG. 65 shows a configuration of an integration management module according to an embodiment of the present invention; FIG. 66 shows a process of generating one applicant name-based patent information service system from entire applicant name patent information database 2300, and generating an inventor name-based patent information service system according to each inventor included in a patent document set included in the applicant name-based patent information service system, according to an embodiment of the present invention;

FIG. 67 shows a process of obtaining an applicant name list, generating applicant name-based patent information service systems corresponding to the number of applicants according to each list from entire applicant name-based patent information database 2300, and generating an inventor name-based patent information service system according to each inventor included in the patent document set included in each of the generated applicant name-based patent information service systems, according to an embodiment of the present invention;

*114FIG. 68 shows a process of obtaining an applicant name list from an obtained document set, generating applicant name-based patent information service systems corresponding to the number of applicants according to each list from entire applicant name-based patent information database 2300, and generating an inventor name-based patent information service system according to each inventor included in the patent document set included in each of the generated applicant name-based patent information service systems, according to an embodiment of the present invention;

FIG. 69 shows an embodiment of a method, which, upon receiving an order requesting the generation of an applicant name-based patent information service system, generates a single applicant name-based patent information service system from the entire applicant name-based patent information database 2300, and generates an inventor name-based patent information service system for each inventor included in a patent document set belonging to the applicant name-based patent information service system;

FIG. 70 is a flowchart showing a method for generating an applicant name-based patent information service system in the unit of a nation, wherein the nature of the generation of the applicant name-based patent information service system in the unit of a nation is to generate the applicant name-based patent information service system and the inventor name-based patent information service system in the unit of a nation from applicant name-based patent information database 2300 in the unit of a nation, and it is thus possible to apply all of the methods as shown in FIGs. 66 to 69;

FIG. 71 shows an embodiment of a process for generating the applicant name-based patent information service system and the inventor name-based patent information service system in the unit of a nation by using family information; FIG. 72 shows an embodiment of a process for generating an applicant name-based patent information service system and an inventor name-based patent information service system in the unit of a nation by using a priority claim number;

FIG. 73 shows an embodiment of a process for generating an applicant name-based patent information service system and an inventor name-based patent information service system in the unit of a nation by using family information!

FIG. 74 shows an embodiment of a process for generating an applicant name-based patent information service system and an inventor name-based patent information service system in the unit of a nation by using family information and priority claim information;

FIG. 75 shows an embodiment of a process performed by both the applicant name-based patent information service system generating engine and the inventor name-based patent information service system generating engine when new documents are added;

FIG. 76 shows a process that an applicant name-based patent information service system administrator performs the CRUD for its managed document set, according to an embodiment of the present invention;

FIG. 77 shows a process that an inventor name-based patent information service system administrator performs the CRUD for its managed document set, according to an embodiment of the present invention;

FIG. 78 illustrates a method of processing the counting by the counting pre-processing module according to an embodiment of the present invention; FIG. 79 illustrates a method of processing the weight by the expense expenditure view weight pre-processing module 3311 according to an embodiment of the present invention;

FIG. 80 illustrates a method of pre-processing the weight by the citation view weight pre-processing module in the citation view is closely similar with that by the expense expenditure view weight pre-processing module 3313 according to an embodiment of the present invention;

FIG. 81 illustrates a method of processing weight by the dispute view weight pre-processing module 3315 according to an embodiment of the present invention; FIG. 82 illustrates a method of processing weight by the concentration rate view weight pre-processing module 3317 according to an embodiment of the present invention;

FIG. 83 illustrates a method of processing the weight by the concentration rate view weight pre-processing module according to an embodiment of the present invention; FIG. 84 illustrates a method of processing weight by the applicant-based weight pre-processing module 3331 according to an embodiment of the present invention!

FIG. 85 illustrates a method of processing weight by the inventor- based weight pre-processing module 3333 according to an embodiment of the present invention;

FIG. 86 illustrates a method of processing weight by the agent- based weight pre-processing module 3335 according to an embodiment of the present invention; FIG. 87 illustrates a method of processing the family information by the family information pre-processing engine of the family information pre-processing module 3810 according to an embodiment of the present invention;

FIG. 88 illustrates a method of citation information pre- processing that represents the citation number according to an embodiment of the present invention;

FIG. 88 illustrates a method of obtaining the forward citation document information that is the information on a forward application document citing the specific document and then store the forward citation document information in the document information of the specific document, according to an embodiment of the present invention;

FIG. 90 illustrates a method for pre-processing citation information according to an embodiment of the present invention;

FIG. 91 illustrates a processing method of a patent classification code pre-processing engine for processing a modified patent classification code according to an embodiment of the present invention;

FIG. 92 illustrates a method of generating a database of the modified patent classification codes by a hierarchical modified patent classification code generating module according to an embodiment of the present invention;

FIG. 93 illustrates a method of generating a USFC tree structure from USPC patent classification codes according to an embodiment of the present invention;

FIG. 94 illustrates a method of allocating USPC patent classification codes to modified patent classification codes corresponding to the USPC tree structure of FIG. 93 according to an embodiment of the present invention;

FIG. 95 illustrates a modified tree structure including the modified patent classification codes according to an embodiment of the present invention, which has the same structure as the USPC tree structure generated from the USPC patent classification codes of FIG. 93;

*FIG. 96 illustrates that the USPC patent classification codes of FIG. 93 can be generated by using an Index to U.S. Patent Classification file (a.k.a., Classification Index file;

*FIG. 97 illustrates a patent classification code pre-processing method according to an embodiment of the present invention;

FIG. 98 illustrates a method of determining a representative applicant name by a representative applicant name pre-processing module according to an embodiment of the present invention!

FIG. 99 illustrates a representative applicant name determining method using a priority number according to an embodiment of the present invention; FIG. 100 is a flowchart showing an example of pre-processing of statistic values according to each patent classification code by the patent classification code-based statistic pre-processing module;

FIG. 101 illustrates a method of generating a statistic value, parameter or calculation value, which automatically includes lower patent classification codes of a given patent classification code according to an embodiment of the present invention!

FIG. 103 illustrates a method of generating a patent classification code-based statistic value, parameter, or calculation value, which reflects backward citation information or forward citation information including lower patent classification codes, for a subset of a particular document set, according to an embodiment of the present invention;

FIG. 104 illustrates a a method of pre-processing and then comparing multiple patent classification codes for at least two document sets by the homogeneous multi-patent classification code correlation pre- processing module according to an embodiment of the present invention;

FIG. 105 illustrates a representing phrase pre-processing method according to an embodiment of the present invention;

FIG. 106 illustrates a representing phrase pre-processing method according to another embodiment of the present invention;

FIG. 107 illustrates a representing phrase pre-processing method according to another embodiment of the present invention;

FIG. 108 illustrates an analysis index obtaining method according to an embodiment of the present invention; FIG. 109 illustrates a method of obtaining a to-be-analyzed patent document set according to an embodiment of the present invention;

FIG. 110 illustrates a patent trend analysis according to an embodiment of the present invention;

FIG. Ill illustrates a simplified analysis according to an embodiment of the present invention;

FIG. 112 illustrates a citation analysis according to an embodiment of the present invention;

FIG. 113 illustrates a cited-by analysis according to an embodiment of the present invention; FIG. 114 illustrates an examination citation analysis according to an embodiment of the present invention;

FIG. 115 illustrates a multiple patent classification code analysis according to an embodiment of the present invention;

FIG. 116 illustrates a multiple patent classification code analysis according to another embodiment of the present invention;

FIG. 117 illustrates a multiple patent classification code analysis according to another embodiment of the present invention!

FIG. 118 illustrates an individual unit patent information system multi-stage grouping module operating method according to another embodiment of the present invention;

FIG. 119 illustrates a patent information system arrangement generating method according to another embodiment of the present invention;

FIG. 120 illustrates a screen implementing an applicant-based patent information system according to an embodiment of the present invention, which shows a list of 500 top Korean applicants according to the order of the number of their applications, wherein Republic of Korea has been selected in the nation tab, and a top 500 tab has been selected in the top 500 of its lower tab, the stock market-listed company, KOSDAQ- listed company, numerously-filed company, entire companies;

FIG. 121 illustrates a screen implementing an applicant-based patent information system according to another embodiment of the present invention, which shows a list of 500 top US applicants according to the order of the number of their applications, wherein US has been selected in the nation tab, and a top 500 tab has been selected in the top 500 of its lower tab, the stock NYSE-Iisted company, NASDAQ-listed company, numerously-filed company, and entire companies;

FIG. 122 illustrates a screen implementing an applicant-based patent information system according to another embodiment of the present invention, which shows a list of 500 top Japanse applicants according to the order of the number of their applications, wherein Japan has been selected in the nation tab, and a top 500 tab has been selected in the top 500 of its lower tab, the JPl-Iisted company, JP2-1isted company, JPM (MOTHERS-listed company), JASDAQ-listed company, and entire companies; FIG. 123 illustrates a screen implementing an applicant-based patent information system according to another embodiment of the present invention, which shows a list of 500 top European applicants according to the order of the number of their applications, wherein European has been selected in the nation tab, and a top 500 tab has been selected in the top 500 of its lower tab, UKl (London stock market), AIM (London stock market), OVERSEAS LISTED, EURONEXT, and entire companies;

FIG. 124 shows one implemented embodiment obtained when a U.S patent tap is selected from a country selection of a patent list of a patent portfolio inside of a patent information system of 3CCM, which appears when 3COM (number 6), which is one of companies registered on the NASDAQ has been selected;

FIG. 125 shows a screen, on which the inventor list of 3CCM is displayed when U.S has been selected on the country tab after clicking the inventor list menu in FIG. 124; FIG. 126 shows a screen, on which a patent document list relating to an inventor is displayed when Aldous Stepha..(numeral 9), has been selected from the inventor list of FIG. 125;

FIG. 127 shows an implemented embodiment about the each-year's application number of each IPC in an IPC sub-class level based on the total number of applications, which is shown when a statistic analysis tab has been selected, a technical area SA(systematic analysis) menu in FIG. 124 has been selected, and U.S in a nation tab has been selected. In this screen, a simplified analysis screen, which is shown when drill-down and a specific cell value has been selected, has been shown as a pop-up; FIG. 128 shows an embodiment showing information about a competitive applicant based on the each-year's application number of each USPC in an USPC sub-class level based on the total number of applications, which is shown when the statistic analysis tab in FIG. 124 has been selected, the competitive company SA (systematic analysis) menu has been selected, and U.S has been selected from the nation tab;

FIG. 129 is a view showing an embodiment of an each-year's analysis result regarding a numerously-citing applicant when a backward citing document set of the present invention is regarded to be a to-be- analyzed patent document set in a case where the entire application document set of 3CCM, which is shown when, respective to 3CCM, the statistic analysis tab has been selected, the U.S citation SA (systematic analysis) menu has been selected, and U.S has been selected from the nation tab in FIG. 124, has been regarded as citation reference document set; FIG. 130 shows an implemented embodiment showing each-year's numerously-filed inventor of each IPC in an IPC sub-class level based on the total number of applications, which is shown when the statistic analysis tab respective to 3CQM has been selected, the inventor SA (systematic analysis) menu has been selected, and U.S has been selected from the nation tab in FIG. 124;

FIG. 131 is a view showing an embodiment of a method for generating processed patent information about a patent document set by automatically including information about low rank patent classification; FIG. 132 is a view showing an embodiment of a method for generating processed patent information including multi-dimensional analysis when a patent document set has been selected;

FIG. 133 is a view showing an embodiment of a method for performing convergence analysis when a patent document set has been selected; FIG. 134 is a view showing an embodiment of a method for providing patent document information corresponding to a value in a case where the value has been selected;

FIG. 135 is a view showing an embodiment of a method for providing multi-dimensional analysis according to each applicant, patent classification code, inventor, and agent;

FIG. 136 is a view showing an embodiment of a method for determining if the patent searched in the first nation can be available in the second nation by using family information;

FIG. 137 is a flowchart showing an embodiment of a method for determining the availability possibility in the second nation, in which JP is designated as the first nation, and KR, US, and EU are designated as the first nations respective to #numeral 1 patent document in the method of FIG. 136;

FIG. 138 shows an embodiment of a search result list view obtained when the methods suggested in FIGs. 136 and 137 have been performed, wherein availability information in an interval phase according to each nation has been included in each item of the search result list;

FIG. 139 shows an embodiment of a monitoring service/directory service method of the present invention, in which a technology tree is configured by using a multi-stage hierarchical structure according to the user, and a large variety of information including a search formula is included in the technology tree;

FIG. 140 illustrates an application of the method of FIG. 139 to a product according to an embodiment of the present invention; FIG. 141 illustrates an application of the method of FIG. 139 to a company system of a competing company according to an embodiment of the present invention;

FIG. 142 illustrates a network diagram obtained by implementing a convergence analysis for inventors according to an embodiment of the present invention! FIG. 143 illustrates a network diagram obtained by implementing a convergence analysis on technological keywords and representing phrase according to an embodiment of the present invention; and

FIGs. 144 to 145 show examples of an interface used for implementing the patent information system of the present invention, in which the interface includes many interfaces relating to option selection. [Mode for Invention] PREMISE INFORMATION

Format of and information included in a disclosed patent document

A patent office of every nation issues gazettes for disclosing applied patents or registered patents satisfying predetermined conditions. Although there are some differences between nations, the disclosed patent information basically includes bibliographical information and information relating to a patent text.

The bibliographical information indispensably includes information on the nation issuing the patent document, a title of the invention, and information on the applicant (including assignee, as used herein)/patentee (including assignee, as used herein), information on the inventor, information on the patent classification code (at least one code selected from IPC, USPC, FI, FT, ECLA codes, as used herein), information on various dates such as the application date, and information on various numbers such as the application number. Further, the bibliographical information selectively includes abstract information, information on a representative drawing when the patent includes at least one drawing, information on reference (which includes at least one of reference information on a prior art disclosed by the applicant, a document searched or examined by the examiner, and information on the patent classification code), information on a priority, information on the examiner, etc. The bibliographical information may further include information on the abstract, and a representative claim or claim 1.

The patent text contains information on technical contents of the patent. A text of a typical patent document includes information on the title of the invention, information on the configuration of the invention, information on the claims, and information on at least one of the effect of the invention, industrial applicability, the technical subject to be solved by the invention, description on drawings, and the prior arts. Further, the patent text inevitably or selectively includes information on drawings according to the type of the inventive technology or selection of the inventor.

Patent classification code information and patent classification code system

Each patent document is assigned at least one patent classification code according to at least one patent classification code system. Each document issued by the nation is such as Republic of Korea,

United States, Japan, and Europe (EPO) is assigned an IPC (International Patent Classification) code. In the case where a nation has its own specific classification system (such as USPC or UPC of the US patent office, F-Term/FT or FI (File Index) of the Japanese patent office, ECLA of the Europe patent office), each patent document is also assigned a patent classification code according to the specific classification system of the corresponding nation issuing the patent document.

According to each patent classification code system, each classification code corresponds one-to-one to a description of the corresponding patent classification code. In each patent classification code system, one dot or dots may be provided to the title information of a classification code below a predetermined depth. The following example shows the hierarchy of the patent classification code system, existence of the title information, existence of dot(s) providing information on a relative position before the title information in the classification system.

Section : H electricity Class : HOl basic electrical element Sub-class : HOlF magnet main group: HOlF 1/00 magnet or magnetic body characterized by its magnetic material 1-dot sub-group :l/01 *those made from inorganic materials

2-dot sub-group -'1/03 **those characterized by coercive force

3-dot sub-group : 1/032 *** those made from hard magnetic material

4-dot sub-group M/04 **** metal or alloy 5-dot sub-group :1/047 ***** alloy characterized by composition

6-dot sub-group :1/053 ******those including rare-earth metal

In the case of the IPC shown above, the patent classification codes corresponding to all the sub-groups have a structure of "sub-class" + "number/number", and at least one dot is attached to each of the titles corresponding to most of these patent classification code (the dot or dots are attached to most of the titles except for the title of number/00, which is usually a code belonging to the main group). The smaller the number of dots attached to a title, the higher the concept of the title relative to its surrounding in the technical classification. The larger the number of dots attached to a title, the lower/the more detailed the concept of the title relative to its surrounding.

As described above, the IPC has a multi-stage hierarchical structure (tree structure). Meanwhile, the USPC, FT, and ECLA also have a multi-stage hierarchical structure. The multi-stage hierarchical structure of each patent classification code can be processed into a database.

Introduction of a patent information system 1 according to the present invention

First, a structure of the patent information system 1 according to the present invention will be briefly described with reference to FIGs. 1 to 13. The patent information system 1 according to the present invention is connected to a user computer 300 through a wire/wireless network 200. The user computer 300 includes a computer used by a human, such as a PC, a notebook, a wire/wireless communication terminal, or a game console, and any computer connected to the patent information system 1, such as a third server other than the patent information system 1, a server of a system or an organ, or a third program module. When the patent information system lprovides a web service, the patent information system lis required to be equipped with a web service supporting module.

The patent information system 1 includes a DB unit 20 relating to various types of data such as patent data, a pre-processing module 30 for various pre-processing, a patent information processing basic module 40 for processing and analysis of patent information, a support module 50 (including a subscriber information processing module 501, a multi^¬ national language processing module 502, and a translation module 503) for various supports such subscriber management and policy management, a patent intelligence module 60 relating a high quality analysis of patent information (since the patent intelligence module usually generates high quality analysis patent information mainly through patent analysis, so it is also named a patent analysis intelligence module or briefly an analysis module, which have the same meaning as used herein), and a hierarchically united patent information service system for generating a patent information service assembly in the unit of an applicant, inventor, or agent.

*226In large, the patent information system 1 includes six elements.

The first element is various pre-processing modules, which perform various pre-processings of an obtained patent document according to purposes thereof.

The second element is the patent information processing basic module 40, which includes modules for performing: 1) search function, 2) analysis function, 3) monitoring function, 4) analysis function, 5) patent document set-obtaining obtainment function, and 6) various multi-stage directory generating function, reporting function.

The third element is the patent analysis intelligence module 60, which generates various types of in-depth patent analysis information including 1) reference analysis and comparison analysis.

The fourth element is a patent information service system generating module, which generates a patent information system 1 according to the applicant, inventor, agent, and patent classification code.

The fifth element is the support element 50, which performs various supporting functions for application of the inventive idea in various environments of the network 200, including 1) subscriber information processing, 2) multi-national language processing, 3) translation processing, 4) web service processing.

The sixth element is various databases, which include:

1) patent document master DB 202,

2) patent classification code master DB 203, 3) various master DBs including subject master DB 204,

4) various master DBs relating to analysis,

5) various support DBs (subscriber DB 206-2, menu DB 206-3, policy DBs including a representing phrase extracting policy DB 206-1-1 and a weight granting policy DB 206-1-2, a patent index DB 206-1-3, an analysis query formula DB 206-1-4, etc.),

6) various secondarily-processed DBs (including text mining DB 207-1 and representing phrase DB 207-2),

7) multiple patent classification code-relating DBs 207-3.

The six elements described above are representative elements. Therefore, it is natural that another construction described in the specification of the present invention is included in the system 1 of the present invention, and there will be no particular comment on functions and modules (such as firewall, subscriber management, etc.) for configuring a computer system, which are obvious to one skilled in the art.

The DB unit 20 includes a DBMS 201 for managing the DBs, a patent document master DB 202 storing synthetically storing patent documents, a patent classification code master DB 203 storing information on the patent classification code, a subject master DB 204 storing information on the applicant, inventor, company, etc., an operation result table DB 205 storing results of multi-dimensional analysis for multi-dimensional analysis, a support DB unit 206 storing information on subscribers and various policies or options, a secondary processing DB unit 207 storing results of processing of various data. The menu DB includes marking methods for the menus according to each nation or each language. The patent information system 1 of the present invention reads locale information in a browser of an accessing user, extracts a menu of a nation/language corresponding to the extracted locale information from the menu DB, generates a screen including the menu made by the extracted language, and transmits the generated screen to the user.

Pre-processing module 30

Now, various pre-processing modules of the present invention are described. The pre-processing modules 30 may include at least one of: 1) a root patent document processing module 301-1 2) a counting pre^¬ processing module 3100 3) a patent classification code-based statistic pre-processing module 3200 4) a weight pre-processing module 3300 5) a citation information pre-processing module 34006) a patent classification code pre-processing module 301-3-1 or 3500 7) a representative applicant pre-processing module 301-4-1-1 or 3600 8) a representing phrase extraction pre-processing module 3700 9) a family information pre^¬ processing module 3800 10) a multiple patent classification code relationship pre-processing module 3900 11) a rightful party change information pre-processing module 302 and 12) an administration processing information obtaining module 303.

Root patent document processing module 301-1

Now, the root patent document processing module 301-1 of the present invention will be described. The root patent document processing module 301-1 processes obtained patent document data (XML, SQML or another type) according to each nation (a patent document of a PCT application is published by the WIPO, which is considered as a nation although it is not a nation). The root patent document processing module 301-1 performs predetermined processings of the obtained patent document data, and then works the obtained patent document data into modified patent document data of a format processible by the patent information system 1 of the present invention. The predetermined processings may include: 1) error filtering 2) patterned error correction 3) manual error correction 4) conversion of SGML document data to XML document data; and 5) standard format conversion. The root patent document processing module 301-1, by passing the obtained patent document data through at least one predetermined error check filter 301-1-1, determines if the obtained patent document data satisfies a format that can be treated by the patent information system 1 of the present invention.

Patent document master DB generation module 301-2 The patent document master DB generation module 301-2 of the present invention generates a patent document master DB 202 based on the patent document data generated by the root patent document processing module 301-1 and information processed by various pre-processing modules of the present invention. Sub-modules included in the patent document master DB generation module 301-2 of the present invention include: 1) a bibliographical DB generation module 301-2-1 for generating a root bibliographical DB 202-1-1, a processed bibliographical DB 202-1-2, a translation bibliographical DB 202-1-3, and a representative drawing DB 202-1-4; 2) a document DB generation module 301-2-2 for generating a patent document DB 202-2-1, a translated patent document DB 202-2-2, and a processed patent document DB 202-2-3; 3) a rightful party change DB generation module 301-2-3 for generating a US assignee change DB 202-3-1 and a right establishment change DB 202-3-2; 4) a family information DB generation module 301-2-4 for generating a nation-based patent status DB 202-4-1 and a nation-based family DB 202-4-2; 5) a citation information DB generation module 301-2-5 for generating an applicant-citation information DB 202-5-1 and an examiner-citation information DB 202-5-2; and 6) an administration information DB generation module 301-2-6 for generating an administration information DB 202-6. The patent document master DB 202 includes a bibliographical DB, a document DB, a right change information DB, a family information DB, a citation information DB, an administration processing information DB.

The bibliographical DB includes root bibliographical information obtainable directly from the patent document data, processed bibliographical information obtained by processing the patent document data itself or the entire patent document data set, translated bibliographical DB from languages, which require translation from among the root bibliographical information and the processed bibliographical information, and other information. The bibliographical DB is generated by the bibliographical DB generation module of the present invention.

The document DB includes patent document data itself, translated document data, a part or all of which has been translated into various languages, and processed patent document data converted from the patent document data or translated patent document data into various types of files in formats of pdf, doc, tiff, html, etc. The patent document data, the translated patent document data, the processed patent document data may exist in the form of a patent document DB, a translated patent document DB, and a processed patent document DB, respectively. The document DB is generated by the document DB generation module of the present invention.

The rightful party change information DB includes an assignee change information DB for the US patent documents, and may include information on names of patentees and patent applicants managed by the patent offices of various nations, and information on establishment registration and cancellation registration of various licensees. The rightful party change information DB is generated by the rightful party change DB generation module of the present invention.

The family information DB includes nation-based family data that includes information relating to the family patent document for the respective nations related to an individual patent document, and nation- based patent status data that includes information on status (current status of a nation-based patent application, whether or not to obtain the right, etc.) in the nations of the respective family patent documents. The family information DB is generated by a family information DB generation module.

The citation information DB includes applicant citation information DB generated based on citation information cited by the applicant and examiner-citation information DB generated based on information on other patent documents cited for examining by the examiner for the specific patent application (mainly, the document cited for denying the novelty and inventive step of the patent application in the procedure of the examination). The citation information DB is generated by the citation information DB generation module.

The administration processing information DB includes data generated based on administration processing information of the respective patent documents generated by the patent offices of the respective nations with respect to the patent documents. The administration processing information DB is generated by the administration information DB generation module. The bibliographical DB generation module obtains processed information of the patent document data and/or entire patent document data processed by various pre-processing modules and generates processed bibliographical data, in which DB including the processed bibliographical data is referred to as the processed bibliographical DB. The processed bibliographical particulars include 1) various counting information, 2) various calculations and evaluations information, etc.

For example, various counting information include 1) the number of applicants and/or patentees, 2) the number of inventors, 3) the number of claims according to each stage, such as a stage of filing an application or registration, 4) the number of specification pages, 5) the number of drawings, 6) the number of kinds of a patent classification code, 7) the number of patent classification codes according to the kind, 8) the number of references(=the number of forward citations)/the number of backward citations, 9) the number of examiner-cited references/the number of backward citations, 10) the number of patent classification codes searched by the examiner, 11) the number of priorities, 12) the number of nation- based family patents including one's nation, 13) the number of total families, 14) the number of independent claims, 15) the number of dependent claims, 16) the number of patents for each nation among the references, 17) the total number of patent documents among the references, 18) the number of non-patent documents among the references, etc. such counting information is processed by the counting pre-processing module of the present invention. The details will be described in a description of the counting pre-processing module. For example, the various operation and evaluations information include 1) an calculation value of various patent indexes, such as the total number of applications, the total number of registrations, occupancy rate, concentration rate, activity rate, etc. with respect to the patent classification code included in the patent document of the applicant (the patent classification code may include the patent classification code itself, the total sub-patent classification codes of the subject patent classification code, the patent classification codes included in every hierarchy, such as a direct higher/next higher patent classification code, or the patent classification up to its main group), 2) a calculation value of various patent indexes for the patent classification code included in the patent document of the inventor of the subject patent document, 3) various analysis indexes, the details of which will be described in the description of various pre-processing modules and analysis modules. Patent classification code master DB generation module 301-3 The pre-processing module of the present invention includes the patent classification code master DB generation module 301-3 and the DB generated by the patent classification code master DB generation module 301-3 includes:

1) root patent classification code DB 203-1 storing raw data of the various patent classification codes obtained from the patent offices of the respective nations or patent root data sources;

2) modified patent classification code DB 203-2 storing various patent classification codes modified suitable for the usage of the patent information system 1 of the present invention, which is generated by a modified patent classification code generation module 301-3-3;

3) total higher patent classification code set DB 203-3 collecting and storing all the higher patent classification codes of any patent classification code, which is generated by a total higher patent classification code set generation module 301-3-1-1; 4) sub-patent classification code set DB 203-4 storing information on direct sub or every sub-patent classification code of any patent classification code, which is generated by a patent classification code set generation module 301-3-1-2, in which the patent classification code set generation module 301-3-1-2 includes a direct sub-patent classification code set generation module 301-3-1-2-1 for generating only the direct sub-patent classification code and a total sub-patent classification code set generation module 301-3-1-2-2 for generating all the sub-patent classification code sets;

*2725) patent classification code tree table DB 203-5 storing a patent classification code system in a tree structure, which is generated by a patent classification code tree table generation module (not shown); and

6) total higher patent classification code table DB 203-6 storing the total higher patent classification code of any patent classification code according to the level, which is generated by a total higher patent classification code set generation module 301-3-1-1. In the meantime, if there is renewal in the patent classification code, the renewal is processed by a patent classification renewal module 301-3-5 of the present invention, and the preset module relating to the renewed patent classification code performs the preset processing.

Hereinafter, the patent classification code will be synthetically described.

Multi-stage attribute of patent classification code The patent classification code having a multi-stage hierarchical structure (tree structure) has a correspondence correlation between the patent classification code and the title information of the patent classification code, a plurality of examples of which has been suggested in the specification of the present invention. A search engine (that is identical to the search module 401 in the present invention) or an extension/wildcard in the DBMS 201 can be used for obtaining information on all the sub-patent classification codes of the given patent classification code in the hierarchical structure. For example, if the extension/wildcard (e.g. ?) is attached to HOlF, to be queried in a search word or query word, the information on all of the sub-patent classifications of HOlF can be obtained. In the same manner, if "HOlFl?" is inputted, it is possible to obtain the information on all the sub- patent classification of HOlFl/00. However, the input of "HOlFl/01?" cannot be allowable for obtaining the information on all the sub-patent classifications of HOlFl/01, because the patent classification codes including the dot(s) provided to the title information of the patent classification code, such as the IPC, have an identical indication pattern (e.g. in the IPC, main group indication!- / + number, in the USPC, class number + number, in the FT, one number + one alphabetic character + number of three places + two alphabetic characters + number, etc.), and indicate a relative parent-child correlation/higher-subordinate correlation with the number of dots. That is, if the dot(s) is not attached to the title information, it is possible to distinguish the parent-child correlation/higher-subordinate correlation only by the indication system of the patent classification code, but if it is, it is impossible to distinguish the parent-child correlation/higher-subordinate correlation only by the indication system of the patent classification code.

In the meantime, it is natural that the patent document corresponding to the higher and sub-patent classification code according to the hierarchical structure attribute corresponds to the higher patent classification code. That is, if the patent classification code of H01F1/04 is assigned to the specific patent document, the patent document corresponds to not only H01F1/032 but also H01F1/03, HOlFl/01, and HOIFI/OO. In interpreting in reverse, the patent information relating to H01F1/03 includes not only patent information relating to H01F1/03, but also the patent information relating to every sub-patent classification code of H01F1/03 according to the hierarchical structure system of the patent classification code. That is, the patent information relating to H01F1/03 must include the patent information relating to not only H01F1/032, H01F1/04, but also every sub-patent classification code of HOlFl/03 according to the patent classification code system.

FIG. 14 shows an example to which such an inventive idea of the present invention is not applied. As shown in FIG. 14, there are more documents corresponding to A61B 3/02 that is a subordinate classification of A61B 3/00, than those corresponding to A61B 3/00.

Three methods for processing of sub-patent classification code Due to the above reason, it is necessary to process the patent classification code including the dot(s) attached to the title information capable of including the information on every sub-patent classification code of the patent classification code, which is processed by the patent classification code pre-processing module 301-3-1 or 3500 included in the patent classification code master DB generation module 301-3 of the invention. It is needed to include the patent information relating to the sub-patent classification code of the given patent classification code in every step of extracting information on the given patent classification code. The representative case of the information extracting step includes 1) search, 2) statistics, 3) analysis, 4) monitor, 5) directory indication, etc. In order to include the patent information relating to every sub- patent classification code of the given patent classification code in the case where the extension (wildcard) cannot be used, it is essentially needed any one processing of the following procedures.

First, the patent classification code system (tree structure) is searched with respect to the given patent classification code by a scheme, such as depth first search, to obtain information on every sub-patent classification code. The information obtaining procedure may perform for every query/search. In the meantime, in the given patent classification code system (tree structure), every sub-patent classification code for every patent classification code or patent classification code including the dot(s) attached to title information is first obtained and stored, and then the information on every stored sub-patent classification code can be obtained upon query/search. Where the data having the tree structure is given, the fact that the respective nodes included in the tree structure is searched by a scheme of depth first search, or the like, for finding a direct sub-node and/or every sub-node of the respective nodes, corresponds to a basic computer science so that the description thereof will be omitted. Further, the information of the obtained direct sub-node and/or every sub-node being corresponded to the specific node to be stored is also the basic concept, so that the description thereof will be omitted.

Second, information on the patent classification code corresponding to its higher patent classification code with respect to every patent classification code is stored according to the stage referring to the patent classification code system (tree structure), and if there is a given patent classification code, it is searched the specific stage in which the given patent classification code is first indicated and is obtained information on every patent classification code included the given patent classification code in the specific stage.

Third, only the marking of the patent classification code is modified while the given patent classification code system (tree structure) is maintained so as to generate the modified patent classification code, and the first and second methods can be performed for the modified patent classification code. Meanwhile, if a scheme of referring the patent classification code system and modifying the patent classification code is appropriately selected, the marking of the patent classification code can be modified capable of responding to range search/range query.

Exemplary description based on H04B 7/00 In the present invention, the tree structure between the patent classification codes is described with an exemplary embodiment of the patent classification code corresponding to main group H04B 7/00 (entitled a wireless transmitting system) based on IPC 7^th edition and its sub-patent classification code, however, such a method and description are not limited to this embodiment, but are identically or equally applied to entire IPC areas. Further, such a method and description are also identically or equally applied to other patent classification code systems (e.g. USPC, FI, FT, or ECLA) in which the dot structure attached to the title indicates the higher and subordinate hierarchy. The main group H04B 7/00, the sub-patent classification code thereof, and the title thereof have the following configuration based on IPC 7^th edition.

H04B 7/00Wireless transmitting system, i.e. Wireless transmitting system using an electromagnetic fields H04B 7/005.control of transmission; equalization H04B 7/01.phase shift reduction H04B 7/015.echo effect reduction H04B 7/02.diversity system

H04B 7/04..those using multiple individual antennas that is separately placed H04B 7/06... in a transmitting station H04B 7/08... in a receiving station

H04B 7/10..those using a single antenna system characterized by polarization or direction attribute H04B 7/12..frequency diversity system H04B 7/14.relay system H04B 7/145..passive relay system

*302H04B 7/15..active relay system H04B 7/155... ground based stations

H04B 7/165 those using angle modulation

H04B 7/17 those using pulse modulation

H04B 7/185...space based or airborne stations H04B 7/19 earth synchronization stations H04B 7/195 asynchronization stations

H04B 7/204... multiple access

H04B 7/208 frequency division multiple access

H04B 7/212 time division multiple access

H04B 7/216 code division or spread-spectrum multiple access H04B 7/22.scattering wave system

H04B 7/24.those for communication between at least two scores H04B 7/26.. at least one score capable of moving The tree hierarchical structure of the above code based on the above dot(s) can be more easily represented as shown in Table 1 below. [Table 1]

At this time, if every higher patent classification code of the given patent classification code is obtained referring to the patent classification code system (tree structure), it is possible to generate a total higher patent classification code table as shown in Table 2. The total higher patent classification code table as shown in Table 2 can be also generated with respect to the USPC, FT, ECLA, FI, etc. as well as the IPC. It is preferred that, when the patent classification codes are renewed, the renewed patent classification codes are obtained and the above table as shown in Table 2 is also renewed. A total higher patent classification code table generation module receives the patent classification code from the patent classification code mater DB so as to generate the higher patent classification code table as shown in Table 2 of the present invention It is preferred that the total higher patent classification code table generation module generates the higher patent classification code table whenever the information on the patent classification code is renewed. The total higher patent classification code table generation module can generate the total higher patent classification code table having an identical formation to the above table, even for the modified patent classification code DB of the present invention.

In Table 2 below, each field, i.e. IPC_ID represents an ID of the specific IPC, S represents a section level, C represents a class level, SC represents a sub-class, MG represents a main group, ldot represents a level in which 1 dot is attached to the title information, 2dots represent a level in which 2 dots are attached to the title information, 3dots represent a level in which 3 dots are attached to the title information (so n dots represent a level in which n dots are attached to the title information), and self represents IPC itself. [Table 2]

That is, in the IPC patent classification code of H04B7/06, when every parent node (the node refers to a diverging score, or the like, in the tree structure) is searched referring to the IPC patent classification code system, the results are H, H04, H04B, H04B7/00, H04B7/02, H04B7/04, H04B7/06, and H04B7/06, and when the results are organized according to the field structure, data represented in IPC_ID69970 can be generated.

When the total higher patent classification code table like Table 2 is provided, every higher patent classification code of the given patent classification code is obtained by 1) searching for the given patent classification code in a self field and obtaining the patent classification code according to each level, while going up to the higher level along a row of the searched patent classification code. The collected and databased total higher patent classification code of the given patent classification code is referred to as total higher patent classification code set DB of the present invention, and every extracted and databased sub-patent classification code of the specific patent classification code by utilizing the total higher patent classification code table is referred to as sub-patent classification code set DB of the present invention. At this time, it is possible to extract only the direct sub-patent classification code of the specific patent classification code by utilizing the total higher patent classification code table, in which the extracted and databased direct sub-patent classification code of such codes is referred to as a direct sub-patent classification code set DB.

When the total higher patent classification code table as shown in Table 2 is provided, a method of obtaining the specific patent classification code and its every sub-patent classification code will be described with an example of H04B7/15. In the first step, if H04B7/15 is searched in the Self field, it is identified that H04B7/15 is present in IPC_ID69976. In the second step, if it is searched which level field corresponds to the patent classification code that is identical to the specific patent classification code, the result is the 2 dots level field. In the third step, if it is searched the patent classification code including H04B7/15 in the 2 dots level field, the searched patent classification codes are H04B7/15, H04B7/155, H04B7/165, H04B7/17, H04B7/185, H04B7/19, H04B7/195, H04B7/204, H04B7/208, H04B7/212, and H04B7/216, which correspond to the specific patent classification code itself and the sub-patent classification codes of H04B7/15. If the specific patent classification code is excluded from the searched patent classification codes, it is of course that every sub-patent classification code can correspond to the specific patent classification code itself. At this time, the information on the specific patent classification code itself and its every sub-patent classification code can be generated as sub-patent classification code set DB. At this time, if every sub-patent classification code of the specific patent classification code is stored, it is possible to generate total sub-patent classification code set DB. Further, it is possible to count the number of every sub-patent classification codes including the specific patent classification code itself (here 11, and 10 sub-patent classification codes excluding the specific patent classification code itself). Preferably, the counting is processed by a counting pre-processing module of the present invention.

Meanwhile, in conceiving a method of extracting only the direct sub-patent classification code of the subject patent classification code, if it is searched the patent classification code in which a level field value of 3 dots is not null (there is the direct subordinate value) and a level field value of 4 dots is null (there is no direct subordinate value) in performing the third step, the resultants are H04B7/155, H04B7/185, and H04B7/204. At this time, if the direct sub-patent classification codes of the specific patent classification code are stored, it can be generated as direct sub-patent classification code set DB. At this time, it is possible to count the number of the direct sub-patent classification codes of the given patent classification code (herein, 3), which is preferably processed by a counting pre-processing module of the present invention. Such the direct sub-patent classification code is required for stage-based open. The stage-based open means to open the IPC directory structure according to the stage (section, class, sub^¬ class, main group, 1 dot sub-group, 2 dot sub-group, 3 dot sub-group, 4 dot sub-group, ... n dot sub-group), in which only the patent classification code included in the corresponding stage is represented. In this case, even though the IPC directory structure is opened according to the stage, the value (search result, statistic value, calculation value, analysis value, etc.) corresponding to the respective stages is preferably the value considering both the specific patent classification code and its every sub-patent classification code. For example, in opening the direct subordinate stage of H04B7/15, even though H04B7/155, H04B7/185, H04B7/204 is included in the next stage of the direct subordinate stage, the represented value of H04B7/155 must include both H04B7/155, and H04B7/165 and H04B7/165 which are subordinate to H04B7/155. Of course, the value of only the specific patent classification code can be represented in the specific case (e.g. user's selection). The afore-mentioned table structure (the table used for searching the sub-patent classification code of the specific patent classification code itself in the three steps) can be utilized for the search described later, obtaining necessary information with an SQL search command, statistic analysis processing, etc., and further if the value of the respective patent classification codes according to the preset rule is previously calculated for analysis convenience, etc., the table structure can be utilized for view, materialized view, (multi-dimensional analysis operation result table) cube, etc. A roll-up operation (any one value includes the values of its subordinate subjects, so as to determine one's value) is utilized in such calculation, and the above table can be utilized for the roll-up operation. For example of the roll-up operation, the year value can be generated by summing up each value of a quarter of a year, and the value of a quarter of a year can be generated by summing up a value of the respective months included in the corresponding quarter of the year. When calculating the value of the specific patent classification code (e.g. the number of applications according to the year), it is possible to generate a value for the specific patent classification code by adding the value of the specific patent classification code and the value of its sub-patent classification code. The details will be described in the analysis of the present invention.

If every IPC code/IPC patent classification code is identically processed by the afore-mentioned method of processing, the table having the above formation can be generated for every IPC patent classification code, and by including the table and the afore-mentioned language-based title information, root IPC patent classification code DB corresponding to IPC among the root patent classification code DB can be generated. In the same method, the table also can be generated with respect to USPC, FT, FI, ECLA, etc.

It is suggested that an implementing example of the information structure showing the child-parent correlation between the patent classification codes with respect to the sub-patent classification code of H04B7/00.

First, it is a result according to a corresponding method (child - > parent) in which a parent patent classification code that is a direct higher classification code of the parent patent classification code corresponds to a child patent classification code, of which the example is shown in Table 3. [Table 3]

Next, an example of the result according to a method (parent -> child) in which the parent patent classification code includes all the child information on the direct sub-patent classification code of the parent patent classification code is shown in Table 4. [Table 4]

Directory generation module 405

The directory generation module 405 of the present invention provides the direct sub-patent classification code according to the patent classification code and indicates the counting value of the patent classification code and direct sub-patent classification code next to the patent classification code and direct sub-patent classification code. The counting value includes at least one selected from 1) the number of every sub-patent classification, 2) the number of direct sub-patent classification, 3) depth of classification, 4) the number of patent documents corresponding to every sub-patent classification, and 5) the number of patent documents corresponding to the direct sub-patent classification, and it is preferred that 1), 2), and 5) are indicated together. The directory generation module (405) reads at least one of 1) to 5) from i) the counting values performed by the counting pre-processing module of the present invention, or from any one of ii) a total quantity analysis multi-dimensional analysis operation result table described later, iii) a patent classification code tree table, or iv) a total higher patent classification code table, so as to indicate 1) to 5).

When the patent classification code information is arranged in a directory form, a sign with color or an icon is indicated in the respective directories so as to improve the accessibility of the user. That is, by providing different colors or preset icon according to the cases where there are many direct sub-patent classifications, there are many sub-patent classifications, the depth is deep, there are many corresponding patent documents, the user can more easily recognize the internal attribute (the number of sub-patent classifications, the depth of the classification, the number of patent documents, etc.) of the directory. Especially, if there are many sub-patent classifications of the specific patent classification code directory or the depth of the classification is deep, it evidence? that there are many patent applications belonging to the corresponding field and the various technologies have been filed so that the specific patent classification code directory is more important than other directory.

FIG. 15 illustrates the directory generation module 405 according to an exemplary embodiment of the present invention. As noted in FIG. 15, from the preset stage of the directory, it is possible to check a box corresponding to at least one patent classification code, and the patent classification code selected with the checked box is searched for each nation or is conducted the patent analysis based on the patent classification code.

The directory generation module 405 of the present invention includes a patent classification code utilization directory generation module 405-1 for generating the directory of the patent classification code corresponding to the multi-stage patent classification code system according to the kinds of at least one patent classification code and a subject-based directory generation module 405-2 for generating a multistage directory according to various analysis subjects, or other confirmable subject which are introduced in the present invention. The operating example of the subject-based directory generation module is shown in FIG. ***. A user or an administrator of the system 10 can generate the directory having the multi-stage hierarchical structure according to the specific subject and make search formula, query formula, a keyword, a patent classification code, etc. to correspond to the directory. The system 10 can search and monitor the patent classification code using the corresponding information and the various analysis of the present specification can be conducted for the searched or monitored result.

Sub-hierarchy patent classification code processing module The sub-hierarchy patent classification code processing module searches a sub-hierarchy patent classification code of the patent classification code obtained from the search formula inputted by the user through the hierarchical patent classification code DB.

At least one from three ways is necessary for including the patent document corresponding to the sub-hierarchy patent classification code of the obtained patent classification code in the search result, which is processed by the sub-hierarchy patent classification code processing module.

First, the sub-hierarchy patent classification code processing module searches every sub-hierarchy patent classification code of the obtained patent classification code based on an obtaining time score using the hierarchical patent classification code DB. The patent classification codes in the hierarchical patent classification code DB include the tree structure, it is possible to search the sub-hierarchy patent classification code. If the specific patent classification code is included in the obtained search formula, the sub-hierarchy patent classification code processing module searches all the sub-hierarchy patent classification codes of the specific patent classification code along the tree structure and transmits the result to a modified search formula generation module. The modified search formula generation module generates new search formula that includes the received patent classification codes and queries the generated search formula to the search engine.

Second, the sub-hierarchy patent classification code processing module searches the sub-hierarchy patent classification codes of every patent classification code, matches the searched sub-hierarchy patent classification codes to a matching table, and stores them in advance, and if the specific patent classification code is included in the obtained search formula, the sub-hierarchy patent classification code processing module directly transmits the sub-hierarchy patent classification codes matched to the specific patent classification code to the modified search formula generation module.

The range of the sub-hierarchy patent classification code extracted by the sub-hierarchy patent classification code extracting module can be classified into two methods. One is a method of extracting the direct sub-patent classification code of the obtained patent classification code and the other is a method of extracting every sub- patent classification code of the obtained patent classification code. At this time, even in extracting the direct sub-patent classification code, the direct sub-patent classification code of the extracted direct sub- patent classification code is extracted, followed by repeating such the extraction, every sub-patent classification code of the obtained patent classification code can be extracted.

For example, if the patent classification code included in the search formula is H04B 7/15, the direct sub-patent classification codes include H04B 7/155, H04B 7/185, and H04B 7/204, and every sub-patent classification codes include H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216. It is well-known to a one skilled in the art that the direct sub- patent classification code and every sub-patent classification code can be easily extracted in both the child -> parent corresponding correlation and the parent -> child corresponding correlation directly or recursively.

The subordinate hierarchy patent classification code extracting module combines and stores the extracted subordinate hierarchy patent classification code. If the subordinate hierarchy patent classification code extracting module stores only the direct sub-patent classification code of the patent classification code of H04B 7/15 included in the search formula, it combines 104B 7/15, H04B 7/155, H04B 7/185, and H04B 7/204^Λ to store, and if subordinate hierarchy patent classification code extracting module stores every sub-patent classification code together with the direct sub-patent classification code of the patent classification code of H04B 7/15 included in the search formula, it combines and stores ^04B 7/15, H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216^Λ.

Modified search formula generation module

The modified search formula generation module unifies the combined patent classification code with the original existing search formula so as to generate the modified search formula.

For example, if the search formula of 1teyword=wireless and active and applicant = Samsung Electronics and IPC = H04B 7/15 is given, the modified search formula generation module searches for IPC H04B 7/15 in the given search formula, receives and combines the direct subordinate or every sub-patent classification code, and generates the modified search formula in which the classification codes are combined, such as ^keyword ^wireless and active and applicant = Samsung Electronics and IPC = H04B 7/15 OR H04B 7/155 OR H04B 7/185 OR H04B 7/204 or ^Λ keyword =wireless and active and applicant = Samsung Electronics and IPC = H04B 7/15 OR H04B 7/155 OR H04B 7/165 OR H04B 7/17 OR H04B 7/185 OR H04B 7/19 OR H04B 7/195 OR H04B 7/204 OR H04B 7/208 OR H04B 7/212 OR H04B 7/216. The following search formula is preferable.

In summarizing the function of the sub-hierarchy patent classification code processing module, when the patent classification code is included in the obtained search formula, the sub-hierarchy patent classification code processing module searches information relating to the sub-hierarchy patent classification code of the patent classification code so as to allow the modified search formula generation module to process the searched information. The search engine receives the search formula including the patent classification code from the computer of the user. At this time, if the patent classification code included in the obtained search formula includes the dot(s) attached to the title information, the search engine obtains the sub-patent classification code in the patent classification code system through the patent classification code pre-processing engine, ties the obtained patent classification code with OR, generates the modified search formula including the patent classification code tied with

OR, and conducts the search with the modified search formula. The respective steps are identical to those described above.

*In the meantime, a typical search engine supports the range search. For example, in searching H04B 7/15, it is possible to receive a response of the search result more quickly when querying to the search engine with "From H04B 7/15 to H04B 7/216", rather than when querying with tieing H04B 7/15 and H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, H04B 7/216 with OR. Especially, if the more the sub-hierarchy patent classification codes, the larger the difference of the search speed. The elements of the modified search formula are slightly differently generated according to whether or not the search engine or DBMS supports the processing of an irregular range symbol for the range. The irregular range symbols processing means to process the case where the symbol written in both ends of the range are different, like ^ΛA≤, E>^Λ. If the search engine supports the irregular range symbol processing, it generates the modified search formula element of ^(patent classification codes ≥ obtained patent classification code, patent classification codes < next sibling of the obtained patent classification code)^Λ with respect to the obtained patent classification code. If the search engine does not support the irregular range symbol processing, the modified search formula element of ^Λ(the obtained classification code) or (patent classification codes ≥ the obtained patent classification code, patent classification codes < a next sibiling of the obtained patent classification code) is generated. If the modified patent classification code arranged in a lexical manner is used, the modified patent classification code is of course processed by the above method.

If the obtained search formula is ^keyword tireless and active and applicant = Samsung Electronics and IPC = H04B 7/15, if the search engine supports the irregular range symbol processing, the search formula element is ^keyword =wireless and active and applicant = Samsung Electronics and IPC >= H04B 7/15 and IPC <H04B 7/22^ but if the search engine does not support the irregular range symbol processing, the search formula element is ^eyword^ireless and active and applicant = Samsung Electronics and IPC - H04B 7/15 or (IPC > H04B 7/15 and IPC <H04B 7/22). The core of the patent classification code pre-processing engine of the present invention is that when the specific patent classification code is given, the patent classification code pre-processing engine automatically processes even the every sub-hierarchy (child) or direct sub-hierarchy patent classification of the specific patent classification code and reflects it to every search formula and query formula for the search, analysis, monitor, etc.

If the search result comes out with the modified search formula, the search result can be displayed by clustering according to the searcher's choice. The clustering means a scheme of tieing the search result and representing the tied search result for each group, which is applied to the present invention as follows. If there is a clustering window representing the clustered structure or the selection of the clustering level is available, the search result of a dot sub-group corresponding the preset or preselected clustering level and a sub-group below the dot sub-group is only represented, and if the searcher selects the subordinate or higher clustering level (the clustering level corresponds to the dot level), the search result corresponding to the selected clustering level is represented.

In describing with an example of the search result relating to H04B 7/00, when the clustering of the 1-dot sub-group level is selected, 7/005, H04B 7/01, H04B 7/015, H04B 7/02, H04B 7/14, H04B 7/22, H04B 7/24 corresponding to the 1-dot sub-group are displayed on the clustering window in the form of the sub-directory folder of H04B 7/00, and the search result shows every patent classification code corresponding to H04B 7/00. At this time, when the searcher selects the H04B 7/14, H04B 7/145, H04B 7/15 which are direct subordinate of the H04B 7/14 are represented in the form of the sub-directory folder, and in this case, the search result shows only the patent classification code corresponding to H04B 7/14 and the sub-patent classification code of the H04B 7/14. If the searcher selects the H04B 7/15 again, H04B 7/155, H04B 7/185, and H04B 7/204 which are subordinate to H04B 7/15 are displayed and the search result shows only the patent classification code corresponding to H04B 7/15 and the sub-patent classification code of H04B 7/15.

At this time, for the searcher's convenience, it is possible to indicate any one of or both of the number of direct sub-patent classification codes of the specific patent classification code and the number of every sub-patent classification code at a side of the respective clustering items (folder names of the directory) of the clustering window. It is also possible to indicate the number of patent documents corresponding to the specific patent classification code and every sub- patent classification code of the specific patent classification code. At this time, both of or any one of the number of the applications and the number of registrations can be indicated. For example, if the patent classification code is H04B 7/15, the direct sub-patent classification code is H04B 7/155, H04B 7/185, and H04B 7/204 so as to indicate ^, or every sub-patent classification code is H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, H04B 7/216 so as to indicate ^Λ1(T. It is preferable to display both because the number of direct sub-patent classification codes is small but the number of sub-patent classification codes under the direct sub-patent classification code is much great. This fact is more revealed in newly- emerging technology or a high-tech technology field having great technology specialization speed or particular classification increase/decrease speed. If only the small number of direct subordinate classifications is indicated, even though the technical field has importance, the user may fail to recognize the importance of the technical field so as to ignore and pass over the corresponding classification.

Family information pre-processing module The patent family refers to the set of document information including the document which are directly related to the specific document in domestic or foreign. Generally, the domestic patent family within one nation includes 1) a divisional application, 2) converted application or dual application, 3) patent application claiming the right of domestic priority (the terms for regulating the patent family range is slightly different depending on the nations, such as a continuation-in-part application, reissue patent, etc. in the US). Generally, the foreign patent family includes 1) the International Patent Application (PCT), 2) the application claiming the right of priority under the Paris Convention, 3) the national-phase patent application of foreign application when 1) or 2) is entered or filed with the patent office of the specific nation, and (1) the divisional application, (2) the converted application or dual application, and (3) the application claiming the right of domestic priority (the regulating condition of the patent family range is slightly different depending on the nations, such as a continuation-in-part application, reissue patent, etc. in US) of 3) application. The construction of the family information pre-processing module for preprocessing the patent family information of the present invention is well illustrated in FIG. 9. A family information pre-processing engine 3810 is required for processing the family information, and various DBs are required for obtaining the family information. For example, the necessary DBs include information on the right of priority under the Paris Convention DB 3671, PAJ DB 3673, KPA DB 3675, Inpadoc DB 3677, other family information DB 3830, etc. Especially, Inpadoc DB 3677 serves an important role.

The family information pre-processing module closely relates to processing and updating the patent document master DB, because 1) new patent documents are continuously, published, 2) most of the patent information is obtained for each nation and new patent documents are continuously published according to the respective nations so that the family information can be changed according to 1) or 2) at any time. That is, it is necessary to renew the family information when obtaining 1) or 2) documents, and the family information on obtaining 1) or 2) documents is renewed by the family information pre-processing module of the present invention. The family information can be found through the specific application number indicated in the original application information and priority information of the existing document or new document.

The method of processing the family information by the family information pre-processing engine of the family information pre-processing module is illustrated in FIG. 87. The family information pre-processing module obtains at least one patent document information (S3720), identifies if the family information, such as priority number, divisional application, converted application, or continuation-in-part application, is present in the patent document information (S3730), co-relates the information of the patent document with an unique document number of the family patent document if the family information is present in the patent document information, so as to store it in patent DB or search index (S3740), and counts and stores the number of nation-based family information of the selectively obtained patent document, or adds the number of family information if the unique document number of the family patent document includes the information on the number of family information (S3750).

The family information processed by the family information preprocessing module is stored in the family information DB of the present invention.

Citation information pre-processing module

The citation information pre-processing module of the present invention includes at least one of a citation information obtaining module 3400-1 for obtaining information relating to citing and cited information renewing module 3400-2 for renewing cited information based on citing information. The citation information obtaining module includes an applicant-cited information obtaining module 3400-1-1 for obtaining the citation-related information from the information cited by the applicant in the applicant's patent document and an examiner-cited information obtaining module 3400-1-2 for obtaining information cited by the examiner in the examination. In the meantime, the applicant-cited information obtaining module 3400-1-1 includes an applicant-cited patent document information obtaining module 3400-1-1-1 and an applicant-cited non-patent document information obtaining module 3400-1-1-2. The applicant-cited patent document information obtaining module 3400-1-1-1 operates when the citing information is the patent information. Further, the examiner-cited information obtaining module 3400-1-2 includes an examiner-cited patent document information obtaining module 3400-1-1-1 and an examiner-cited non-patent document information obtaining module 3400-1-2. The citation information pre-processing module processes the citation information obtained by the citation information obtaining module with a preset manner.

The patent citation information includes 1) prior art citation information that is other patent document cited by the applicant as a prior art, 2) reference citation information that is a patent document cited by the examiner as the reference in examining the specific patent document, and 3) indirect citation information. It is natural that the citation is incurred in domestic and foreign. Generally, the citation of other patent as the prior art is included in the prior art document information in the patent document of Republic of Korea and Japan, and is included in the reference information in the patent document of US. Such the prior art citation information can be obtained from the bibliographical particular or body contents of the patent document. Further, many nations publishes the reference citation information, which also can be identified in administration information published from the patent office. The present specification describes with mainly focusing the prior art citation information, but if the obtained reference citation information can be treated like the prior art citation information, the above description can be equally applied thereto. The indirect citation includes indirect backward citation and indirect forward citation. When at least one patent document set is given for processing, the indirect backward citation document set of the patent document set includes the patent documents having the application date that is earlier than that of the to-be-processed patent document and simultaneously citing the patent document that is cited by the to-be-processed patent document. The indirect forward citation document set of the patent document set includes the patent documents having the application date that is later than that of the to-be-processed patent document and simultaneously cited in the patent document that cites the to-be-processed patent document. The indirect citation is resulted from the difference between the patent application date and the patent publication date.

The citation information pre-processing module of the present invention pre-processes the prior art citation information and reference citation information, and as shown in FIG. 5, the citation information pre-processing module includes a prior art citation information pre^¬ processing module for pre-processing the prior art citation information and a reference citation information pre-processing module for preprocessing the reference citation information. Typically, the backward citation is included in the bibliographical particular or body contents of the specific patent document so that the document information relating to the backward citation is stored in the search index of the specific document or specific document information. However, the forward citation relating to the specific document is not stored in the specific document information, but is stored in the search index of the document that cites the specific document or citing document information. Therefore, it is necessary to search the forward citation relating to the specific document, correlating the searched forward citation with the specific document, and store it, which are processed by the prior art citation information pre^¬ processing module.

Prior art citation information pre-processing module The prior art citation information pre-processing module queries the application number of the specific document and the unique number value of the specific document to the search index or DBMS so as to obtain the forward citation information of the specific document that is cited-by. The prior art citation information pre-processing module can store the forward citation information of the obtained specific document in the search index in the step of indexing the specific document and stores it as the forward citation information of the specific document in the patent document master DB. There is few change in the backward citation information (that is written by the applicant on filing the patent application, and there may be addition of the citation, but conventionally the addition of the forward citation rarely occurs) and the forward citation information is the information which must be always monitored when the patent information is renewed. Therefore, when the new patent document is published so the search index is renewed or the patent document master DB is renewed, the prior art citation information preprocessing module queries if the renewed patent document citing the specific document is present so as to renew the forward citation information.

A method of citation information pre-processing that represents the citation number is illustrated in FIG. 88 in detail. The citation information pre-processing module receives at least one citation number present in the specific document (S3820), obtains the citation number in which the attribute of the citation number is not preset (S3830), queries the obtained citation number to the search engine or DBMS so as to receive the citation number having the preset attribute (S3840), and represents the received citation number with the obtained citation number having the preset attribute (S3850). It is necessary to obtain the forward citation document information that is the information on a forward application document citing the specific document and then store the forward citation document information in the document information of the specific document, an exemplary method of which is illustrated in FIG. 89. The citation information pre-processing module obtains specific document data (S3920), obtains at least one of the application number, registration number, and publication number included in the specific document data (S3940), queries the obtained number to a citation field of the search engine or queries the obtained number to the DBMS by limiting the citation field (S3950), receives the forward citation information as the query result (S3960), and stores the forward citation information in the bibliographical particular of the document data (S3970).

Reference citation information pre-processing module The reference citation information pre-processing module pre- processes the reference citation information by the same manner as the prior art citation information pre-processing module pre-processes the prior art citation information.

Counting pre-processing module Hereinafter, the counting pre-processing module will be described. The counting pre-processing module includes an individual document-based counting pre-processing module for processing the counting according to each individual document, and a multiple document-based counting pre~ processing module for processing the counting in the unit of at least two documents.

Individual document-based counting pre-processing module

The individual document-based counting pre-processing module can calculate numerical information of at least one from 1) the number of applicants and/or patentees, 2) the number of inventors, 3) the number of claims for each step, such as the step of filing an application or registration, 4) the number of specification pages, 5) the number of drawings, 6) the number of kinds of a patent classification code, 7) the number of patent classification codes according to the respective kinds, 8) the number of references, 9) the number of references cited by the examiner, 10) the number of patent classification codes searched by the examiner, 11) the number of priorities, 12) the number of nation-based family patents, 13) the number of total families, 14) the number of independent claims, 15) the number of dependent claims, 16) the number of patents for each nation among the references, 17) the total number of patent documents among the references, and 18) the number of non-patent documents among the references.

Further, the individual document-based counting pre-processing module can calculate any one of information on 1) the term from the application date to the registration date, 2) the term from the priority date to the application date, 3) the term from the application date to the date of requesting an examination, and 4) the term from the date of the Office Action to the date of filing a response to the Office Action from the individual patent document or the administration processing information of the patent office that publishes the individual patent document .

Further, the individual document-based counting pre-processing module can obtain information on whether or not to utilize or apply the specific system according to the individual nation from the individual patent document or administration information, in which the information includes 1) whether to utilize the examination request system, 2) whether to utilize the divisional application, 3) whether to utilize the continuation-in-part application, 4) whether to utilize the reissue patent system, 5) whether to request the accelerated examination, 6) whether to utilize the decision appealing trial, 6) whether to occur a non-party litigation. If the system is utilized or applied, the counting value can be set as 1, and if it is not, the counting value can be set as O.

Further, the individual document-based counting pre-processing module searches the patent document master DB relating to the individual patent document (using the search engine or DBMS) so as to obtain and count the information relating to the individual patent document of at least one second nation, in which the information includes 1) the number of nations of application, 2) the number of nations of registration, 3) whether or not to file the PCT application, 4) the number of families according to the nation, and 5) the total number of families.

Further, the individual document-based counting pre-processing module can recognize and count any one of 1) forward citation (the domestic and foreign patent documents citing the specific patent document), 2) backward citation (the previous patent documents cited by the specific patent document), and 3) indirect citation through the search engine and DBMS.

Further, the individual document-based counting pre-processing module can obtain information on 1) whether to request an location, 2) whether or not provide the information, and 3) whether to occur a party trial for the individual patent document. Information 1) to 3) can be obtained directly or by a search method from the Korean Patent Office or the Industrial Attribute Tribunal in Korea.

It is preferred that the value counted by the individual document- based counting pre-processing module is stored in the processing bibliographical DB of the patent information master DB. Further, it is more preferable that the counting value is included in the index that is generated when the search engine indexes the patent document for storing. A method of including the counting value in the index includes at least one of 1) operating the individual document-based counting pre-processing module when the search engine indexes the patent document so as to obtain the counting value, or 2) obtaining the counting value stored in the patent information master DB.

The counting pre-processing module counts 1) the number of total sub-patent classification codes and 2) the number of direct sub-patent classification code of each patent classification code referring to the patent classification code master DB and stores the counting value in the patent classification code master DB.

A method of processing the counting by the counting pre-processing module is well illustrated in FIG. 78. First, the counting pre-processing module obtains at least one patent document information S2820 and calls at least one individual document-based counting pre-processing module according to the obtained patent document S2830, the individual document- based counting pre-processing module processes the counting of the patent document information S2840 and stores the counting processing result in the patent DB or search index together with the document unique number of the patent document information S2850.

Multiple document-based counting pre-processing module Hereinafter, the multiple document-based counting pre-processing module of obtaining a score calculated in the unit of the multiple documents will be described. The multiple document-based counting pre^¬ processing module obtains the score by collecting information on the multiple documents for the subject set counted by the individual document- based counting pre-processing module. That is, the multiple document-based counting pre-processing module generates the counting data for at least two documents set.

*402Weight pre-processing module

Document-based weight pre-processing module

It is impossible to measure weight of the inventive idea included in one patent document by a system. However, it is unreasonable to identically treat the weight of every patent document due to the reason of failing to measure the accurate weight. Therefore, the present invention suggests a method of granting the weight of the patent document on a basis of the information capable of being recognized with the data by the system. The weight of the patent document is granted by a document-based weight pre-processing module. The weight pre-processing module of the present invention includes the document-based weight pre-processing module 3310 and a subject-based weight pre-processing module 3330. The document-based weight pre^¬ processing module 3310 includes an expense expenditure view weight pre^¬ processing module 3311, a citation view weight pre-processing module 3313, a dispute view weight pre-processing module 3315, and/or concentration rate view weight pre-processing module 3317. Further, the subject-based weight pre-processing module 3330 includes an applicant-based weight pre^¬ processing module 3331, an inventor-based weight pre-processing module 3333, and/or an agent-based weight pre-processing module 3335.

*407Expense expenditure view weight pre-processing module From a probable, statistical, or social view, there is high possibility in that the more one expends the expense for the patent, the more the patent is important. There is a high possibility of establishing such an assumption in an identical subject in view of the reasonable expense distribution, and also of establishing such an assumption between the different subjects they are in competition. Investigating the structure of a single patent requiring the expense expenditure, the following weight granting elements can be considered:

1) in the quantity view, (1) the number of claims, the number of independent claims, the quantity of specification (the number of pages), the number of families, and the number of patent classification codes, 2) in the status view, the pending patent application, the registered patent, rejected patent, or abandoned patent, 3) in the procedure view, whether to request an examination, whether to utilize the accelerated examination, whether to request a trial, whether to claim the right of domestic priority, or whether to file a divisional application, 4) in the subject view, the number of applicants or the number of inventors, and 5) in the foreign view, whether to utilize the PCT application, the number of nations entering the foreign national phase. These weight granting elements can influence on the weight of the patent document, and can be obtained from the bibliographical particular information, specification contents information, or administration processing information of the individual document.

In general, if there are many numbers of claims or large quantity of specification, there may be a high possibility in that the contents of the invention is broad and various and it spent lots of time and expenses for the invention, and also the agents may charge larger expense. Further, the more the number of families or the patent classification codes, the more expansive the invention. Further, there may be a possibility in that the registered patent has higher weight than the pending patent application and the patent application requesting the examination or accelerated examination is more important than the non-examination- requested application. If there are two or more applicants, the patent application is a joint application and is a result of cooperation between two or more different subjects, so that the joint application is more important than a non-joint application in a statistic view, and also if there are multiple inventors, the invention has been completed with the cooperative research of the several inventors so as to be more important in the statistic view. Further, generally, the foreign application requires greatly larger expense than the domestic application, so if there is an application in a foreign nation or many foreign applications, the applicant expends the substantial expense for the foreign patents so that it is reasonable to consider that the patent is relatively more important to the applicant.

From the above views, it is possible to recognize the existence and numerical value of the weight granting elements with respect to one document (In view of the problem of existence, if the weight granting elements do not exist, the numerical value is 0, and if the weight granting elements exist, the numerical value is 1, and if the weight granting elements are the number of claims, the numerical value is 1 or higher.) and a weight score is granted according to the weight granting element according to the existence and/or numerical value. The expense expenditure view weight pre-processing module recognizes the existence and numerical value of the particular weight granting elements considered in the respective views 1) to 5) and grants the weight score according to the weight granting elements.

A method of processing the weight by the expense expenditure view weight pre-processing module is well illustrated in FIG. 79. The expense expenditure view weight pre-processing module obtains at least one patent document information S2920, obtains a counting result value according to at least one preset counting references relating to the expense expenditure from the obtained patent document information using the counting pre-processing module or through the patent DB or search index S2930, and obtains the weight according to the counting references of the patent information from the patent information processing policy DB S2940. At this time, the information according to the preset counting reference is referred in the patent information processing policy stored in the patent information processing policy DB. For example, in view of the claim of the patent information processing policy DB according to the expense expenditure view, the patent information processing policy DB stores the policy for processing the weight, such as (the number of claims 0.1 + the number of independent claims 0.3), according to the respective subjects in the respective views. The expense expenditure view preprocessing module generates an expense weight value with counting reference-based weight and the counting result value according to the counting references S2940 and stores the generated weight value in the patent DB or search index together with the unique document number of the patent document information S2950.

*413

Citation view weight pre-processing module

There is high possibility in that the patent cited by a lot of patents may be important patent, like a paper. A module for processing weight of the individual document in the citation view refers to the citation view weight pre-processing module. In classifying the citation view in detail, the weight granting elements includes 1) in the forward citation view, the number of total forward citations, the number of depths of forward citation, the number of first-level forward citations, the average term of forward citation, and the concentration rate of forward citation in comparison with the term and 2) in the examiner citation view, the examiner citation. That is, the more the citation of one's patent directly or indirectly (the patent citing one's patent is cited by other patent), the more the first-level citation that directly cites the specific document, as the average term of forward citation belongs within an appropriate range, and the more the people who cite the subject patent are concentrated on a recent time, the more the important of the subject patent. Further, the patent cited by the examiner in the examination may be more important than the patent that is not cited by the examiner. A method of pre-processing the weight by the citation view weight preprocessing module in the citation view is closely similar with that by the expense expenditure view weight pre-processing module, which is well illustrated in FIG. 80. The citation view weight pre-processing module obtains at least one patent document information (S3020), obtains the counting result value according to at least one preset counting reference relating to the backward citation and/or forward citation, or examination citation from the obtained patent document information using the counting pre-processing module or through the patent DB or search index (S3030), refers the patent information processing policy DB and obtains the weight according to the counting reference (S3040), generates the citation view weight value with the counting-reference-based weight and the counting result value according to the counting reference (S3050), and stores the generated weight value in the patent DB or search index together with the unique document number of the patent document information (S3060). At this time, the information according to the preset counting reference relating to the citation refers to the patent information processing policy stored in the patent information processing policy DB. For example, the patent information processing policy DB in the citation view stores the policy of processing the weight, such as "the number of forward citation countings 0.3", according to the subjects in the respective views. Dispute view weight pre-processing module There is a high possibility in that the more the dispute relating to one patent document, the more important the patent. The disputative elements recognized in the patent information include 1) invalidation trial, 2) location, 3) information provision, 4) passive or active right scope identification trial, etc. The dispute view weight pre-processing module recognizes the existence and numerical value of the weight granting element and grants the weight score according to the weight granting element in the disputative view.

A method of processing weight by the dispute view weight preprocessing module is illustrated in FIG. 81, which is closely similar with that by the expense expenditure view weight pre-processing module and citation view weight pre-processing module.

*The dispute view weight pre-processing module obtains at least one patent document information (S3120), obtains the counting result value according to at least one preset counting reference relating to the dispute from the obtained patent document information using the counting pre-processing module or through the patent DB or search index (S3130), refers to the patent information processing policy DB and obtains the weight according to the counting reference (S3140), generates the dispute view weight value with the counting reference-based weight and the counting result value according to the counting reference (S3150), and stores the generated weight value in the patent DB or search index together with the unique document number of the patent document information (S3160).

That is, the essence of the method of processing the weight by the dispute view weight pre-processing module is to perform the counting of the number of disputes, obtains the weight corresponding to the number of countings from the patent information processing policy DB, and generates the dispute view weight value.

Concentration rate view weight pre-processing module

There is a high possibility in that if the several subjects are concentrated in the specific technical field based on the specific time score, the technical field may be more important. It is rare that one applicant files the applications evenly in the several technical fields, and the plurality of applications are filed for the technical field which the applicant considers important and to which the technical investment is concentrated. Meanwhile, a small number of applications is defensively filed for the technical field which the applicant considers less important. Therefore, if the multiple applicants files the applications for the technical field to which one application belongs so that 1) the number of applications is large, 2) the increase/decrease rate or speed of the application is high, 3) the number of applicants is large, or 4) the variation of the occupancy rate exceeds the preset range in relative with another technical field, there is a high possibility in that the technical field may be more important. The technical field to which the multiple applicants are concentrated can be measured by an analysis element, such as technology attractiveness, the application increase/decrease rate, the application increase/decrease speed, the variation of the occupancy rate, etc. At this time, it is preferred that the technical field of this paragraph is identified from the hierarchical patent classification code included in the patent document, such as IPC, USPC, FI, FT, and ECLA, is recognized in the respective hierarchy units. Of course, it is more preferable that the sub-patent classification code is automatically included and the respective analysis element-based numerical value is calculated. That is, if H04B 7/26 as the IPC is given, it is possible to calculate the various analysis element-based numerical values for the IPC of H04B 7/26 and also calculate the various analysis element-based numerical values in the H04B 7/00 view. At this time, it is more preferable to calculate the various analysis element-based numerical values with including another patent classification codes included in the hierarchically sub-patent classification code of H04B 7/26. The concentration rate view weight pre-processing module of the present invention recognizes the existence and the numerical value of the weight granting element and grants the weight score according to the respective weight granting elements in the concentration rate view.

The method of processing weight by the concentration rate view weight pre-processing module is closely similar with that by the expense expenditure weight pre-processing module and citation view weight pre^¬ processing module. That is, the concentration rate view weight pre- processing module measures the value of the patent information analysis index meaning the concentration rate, such as concentration rate/activity rate/occupancy rate in the technical field of the applicant/patentee with the bibliographical information of the respective individual documents, obtains the weight of the measured value from the patent information processing policy DB, and generates the concentration rate view weight value.

The method of processing weight by the concentration rate view weight pre-processing module is illustrated in FIG. 82, which is closely similar with that by the expense expenditure weight pre-processing module and citation view weight pre-processing module.

The concentration rate view weight pre-processing module obtains at least one patent document information (S3220), obtains applicant information and patent technology classification information from the patent document DB (S3230), calculates, by means of the value of the obtained applicant information and patent technology classification information value, the value of applicant's concentration rate/activity rate/occupancy rate in the patent technology classification or at least one higher patent technology classification code obtained in the patent technology classification code system referring to the patent technology classification code DB on a basis of at least one reference (S3240), obtains the weight information on the reference-based concentration rate/activity rate/occupancy rate from the patent information processing policy DB (S3250), generates a concentration rate view weight value with the weight according to the reference and the concentration rate result value (S3260), and stores the generated weight value in the patent DB and search index together with the unique document number of the patent document information (S3270).

The concentration rate view weight pre-processing module can process the weight based on the inventor, not based on the applicant, and a method of processing the weight by the concentration rate view weight pre-processing module is illustrated in FIG. 83. The concentration rate view weight pre-processing module obtains at least one patent document information (S3320), obtains inventor information and patent technology classification information from the patent document information (S3330), calculates the concentration rate of the inventor for the patent technology classification or at least one higher patent technology classification code obtained in the patent technology classification code system with reference to the patent technology classification code DB with the value of obtained inventor information and patent technology classification information value on a basis of at least one reference (S3340), obtains the weight information on the reference-based concentration rate (S3350), generates the concentration rate view weight value with the weight according to the reference and the concentration rate result value (S3360), and stores the generated weight value in the patent DB and search index together with the unique document number of the patent document information (S3370).

Subject-based weight pre-processing module

If one patent document is given and a subject involved in the patent document is important, there is a high possibility in that the patent filed by the subject may be important. At this time, it is a keypoint to determine which subject is more important. The subject involved in the single patent document includes an applicant, an inventor, and an agent. The subject-based weight pre-processing module pre-processes the weight of the patent document in the subject view. AS shown in FIG. 4, the subject-based weight pre-processing module includes an applicant-based weight pre-processing module, an inventor-based weight pre-processing module, and an agent-based weight pre-processing module according to the subject.

Applicant-based weight pre-processing module If the applicant of the document concentrates on the technical field included in the specific document, there is a high possibility that the application is important. The technical filed on which the applicant of the specific document concentrates can be measured with the analysis element, such as concentration rate, occupancy rate, and AI in the technical field. The applicant-based weight pre-processing module of the present invention recognizes the numerical value of the weight granting element and grants the weight score according to the respective weight granting elements in the analysis element view. A method of processing weight by the applicant-based weight pre^¬ processing module is illustrated in FIG. 84. The applicant-based weight pre-processing module obtains reference information for processing weight based on the applicant (S3420), obtains a result value of the document in the name of the applicant conforming with the reference calculated by at least one document-based weight pre-processing module (S3430), obtains weight information according to the document-based weight pre-processing module from the patent information processing policy DB (S3440), generates a reference-based applicant-based weight value obtained with the respective weight and respective result values of the document-based weight pre-processing module (S3450), and stores the generated weight value in the patent DB or search index together with the unique document number of the patent document information, together with the reference or independently (S3460).

Inventor-based weight pre-processing module

If the inventor of the document concentrates on the technical field included in the specific document, there is a high possibility that the application is important. The technical filed on which the inventor of the specific document concentrates can be measured with the analysis element, such as concentration rate, occupancy rate, and AI (in the formula, the inventor is substituted for the applicant in the applicant- reference formula) on a basis of the inventor of the technical field. The inventor-based weight pre-processing module of the present invention recognizes the numerical value of the weight granting element and grants the weight score according to the respective weight granting elements in the analysis element view.

A method of processing weight by the inventor-based weight pre^¬ processing module is illustrated in FIG. 85. The inventor-based weight pre-processing module obtains reference information for processing weight based on the inventor (S3520), obtains a result value of the document in the name of the inventor conforming with the reference calculated by at least one document-based weight pre-processing module (S3530), obtains weight information according to the document-based weight pre-processing module from the patent information processing policy DB (S3540), generates a reference-based inventor-based weight value obtained with the weight of the respective document-based weight pre-processing modules and the result values (S3550), and stores the generated weight value in the patent DB or search index together with the unique document number of the patent document information, together with or independent of the reference. The weight information may include the policy, such as "registration rate in comparison with total previous applications^χθ.1 + the number of foreign application families^χθ.5".

Agent-based weight pre-processing module If the agent of the document concentrates on the technical field included in the specific document, there is a high possibility that the specification written by the above agent is more in-depth than that written by another agent. The technical filed on which the agent of the specific document concentrate can be measured with the analysis element, such as concentration rate, occupancy rate, and AI (in the formula, the agent is substituted for the applicant in the applicant-reference formula) on a basis of the agent of the technical field. The agent-based weight pre-processing module of the present invention recognizes the numerical value of the weight granting element and grants the weight score according to the respective weight granting elements in the analysis element view. The weight based on the agent is less important that that of the weight based on the applicant or based on the inventor.

A method of processing weight by the agent-based weight pre^¬ processing module is illustrated in FIG. 86. The agent-based weight pre- processing module obtains reference information for processing weight based on the agent (S3620), obtains a result value of the document in the name of the agent corresponding to the reference calculated by at least one document-based weight pre-processing module (S3630), obtains weight information according to the document-based weight pre-processing module from the patent information processing policy DB (S3640), generates a reference-based agent-based weight value obtained with the weight of the respective document-based weight pre-processing modules and the result values (S3650), and stores the generated weight value in the patent DB or search index together with the unique document number of the patent document information, together with or independent of the reference (S3660). The weight information may include the policy, such as "registration rate in comparison with applications*0.3".

The document-based weight pre-processing module and subject-based weight pre-processing module of the present invention grants the preset weight to the weight granting elements obtained from the at least one weight pre-processing modules and calculates the weight in the unit of the document. At this time, the respective weight granting modules according to the respective views or the weight granting modules according to the subjects grant the preset weight to the weight granting element obtained by themselves and calculate the weight in the unit of the document in themselves view. Of course, the respective weights can be variously granted according to the input of the administrator or user.

Subject master DB generation module 301-4 Kind of subject

The subject in the patent document generally includes the 1) applicant, 2) inventor, and 3) agent. The applicant includes an organization, such as a company, and an individual. The DB storing and managing the subject in the present invention is referred to as subject master DB 204, and the module of generating the subject master DB 204 is referred to as the subject master DB generation module 301-4.

The subject master DB 204 further includes represented applicant's name DB, represented inventor's name DB, and represented agent's name DB, and independently further includes company information DB, which will be hereinafter described.

Construction of the subject master DB generation module 301-4

The subject master DB generation module 301-4 of the present invention includes a name-representation pre-processing module 301-4-1. The name-representation pre-processing module includes at least one of an applicant's name representation pre-processing module 301-4-1-1 for processing the representation of the applicant's name, an inventor's name representation pre-processing module 301-4-1-2 for processing the representation of the inventor's name, and an agent's name representation pre-processing module 301-4-1-3 for processing the representation of the agent's name. In the meantime, the subject master DB generation module 301-4 of the present invention further includes a company information DB generation module. The company information DB includes at least one of financial information of the company, company status information, accounts information, a main product, a name of a representative, a company registration number, a taxpayer identification number, a web-site, a telephone number, and a fax number. The subject master DB 204 particularly further includes company information DB 204-1, represented applicant's name DB 204-2, represented inventor's name DB 204-3, represented agent's name DB 204-4, and/or organization information DB 204-5.

Representing phrase extraction pre-processing module

Concept of representing phrase

Next, the representing phrase extracting module of the present invention will be described. First, the concept of the representing phrase will be described. The representing phrase refers to a word or a phrase including at least two words capable of representing the entire specific document or a partial of the specific document. The basic attribute of the representing phrase of the present invention is to substantially rarely come out in the preset representing phrase extraction entire documents set, but often come out by the preset reference or more in the preset given representing phrase extraction entire document set

(including one representing phrase extraction target document) or a part of the representing phrase extraction target document or a common part of the respective documents included in the preset representing phrase extraction target document set.

*457For example, a kind of preset representing phrase extraction entire document set includes 1) a patent document set generated by a method of specifying a patent document set with the specific patent classification code in the specific classification code system in the specific nation DB, 2) a patent document set generated by a method of specifying a patent document set with the specific applicant's name in the specific nation DB, 3) a patent document set generated by a method of specifying a patent document set with the specific inventor's name (i.e. the applicant's name and inventor's name) included in the patent document of the specific applicant in the specific nation DB, 4) a patent document set generated by a method of specifying a patent document set with the specific agent's name in the specific nation DB, 5) a patent document set generated by a method of specifying a patent document set with the specific patent classification code and specific applicant's name in the specific patent classification code system in the specific nation DB, 6) a patent document set generated by a method of specifying a patent document set with the specific patent classification code, specific applicant's name, and specific inventor's name in the specific patent classification code system in the specific nation DB, 7) a patent document set generated by a method of specifying a patent document set with the specific applicant's and specific agent's in the specific nation DB, 8) every patent document set of the specific nation, 9) every patent document set of at least two nations, and 10) a patent document set in which 1) to 9) are specified in the unit of the preset term.

In the meantime, the representing phrase extraction target document set can be any subset of the representing phrase extraction entire document set (a subset including the documents in which at least one specific attribute of one document (e.g. the specific classification code, such as IPC, applicant, term, nation, agent or inventor, or at least one combination thereof) is common).

The representing phrase extracting target document set (a super set including a target set) is a concept opposite to the representing phrase extraction entire document set (target set), and if it is a subset of the representing phrase extraction entire document set, it is sufficient, but the smaller the size of the subset, the better. For example, the representing phrase extraction target document set includes 1) a single specific application, every application of the specific company, an application of the specific inventor, an application of the specific IPC of the specific inventor, every application of the specific IPC of the specific company, every application of the specific year of the specific IPC, etc.

In the meantime, for example, a part of the representing phrase extraction target document includes the claims, the independent claims of the claims, the dependent claims of the claims, the effect of the invention or industrial applicability, and abstract. Meanwhile, the common part of the respective documents of the preset representing phrase extraction target document set can be the aforementioned claims of every document included in the representing phrase extraction target document set.

The representing phrase includes 1) a phrase constructed with only noun, 2) a phrase constructed with a noun and verb, and 3) a phrase constructed with even an adjective and adverb. It is preferred that the length of the representing phrase is equivalent to that of 1 to 5 words with a reference of the number of words, and more preferably 2 to 3 words. Because, the technical concept is typically generated with one word, but in the many cases, the technical concept, technical action, or technical effect is constructed with 2 to 3 words. Next, the step of extracting the representing phrase will be described in detail. The representing phrase extraction pre-processing module extracts the representing phrase. The representing phrase extraction pre-processing module of the present invention includes 1) a language-based morpheme analysis engine (analyzer), 2) a phrase generating engine, 3) a phrase-based counting engine, 4) various dictionaries DB including a thesaurus dictionary/patent technology term dictionary, etc., and 5) a representing phrase extracting engine. The construction of the representing phrase extraction pre-processing module is well illustrated in FIG. 8. The representing phrase extraction pre-processing module includes a representing phrase extracting pre-processing engine 3710 including an algorithm of extracting the representing phrase, representing phrase extraction policy DB 3730 including information on a policy serving as a reference for extracting the representing phrase, and representing phrase-frequency-application number correspondence DB 3750 including information on which location (field, etc.) of which document includes various counting or calculation information, such as a frequency value of the representing phrase, and the representing phrase when the representing phrase is extracted. In the meantime, the representing phrase extraction pre-processing module further includes thesaurus DB 3770 for processing a synonym and representing phrase translation DB 3790 including translation information on the representing phrase using a translation system for processing at least two kinds of languages or a pre-translated dictionary. Hereinafter, the representing phrase extracting pre-processing engine of the present invention will be described in more detail. The representing phrase extracting pre-processing engine extracts a terminology when a sentence or a paragraph of the specific field is inputted.

The phrase generating engine of the present invention generates a phrase array including 1 to 5 words (preferably, 2 to 3 words). For example, it is assumed that the words outputted when one sentence included in one patent document is inputted are a, b, c, d, e, f, g, in sequence. At this time, if a phrase generating unit of the phrase generating engine includes 3 words, the phrase generating engine generates the word-based phrase of abc, bed, cde, def, and efg. The phrase generating engine generates the phrase in the unit of the sentence for the given document or a part of the document. For example of generating the words for the part of the document, the phrase can be generated with respect to the sentences included in the claims. A method of generating the phrase array includes 1) a method of, in the unit of one sentence, if n number of phrases is extracted from a single sentence, generating the phrase of nC2, nC3, nC4, nC5 by means of n number of phrases (e.g. if nC2 is applied, the phrases are generated as ab, ac, ad, ae, af, ag, be, ...fg), 2) if there is m number of phrases in the unit of the document, applying above 1) method for the unit of the document, and 3) applying the phrase including 2 to 5 words in sequence (e.g. if two words are sequentially arrayed, then the phrase array becomes ab, be, cd, de, ef, fg, and if three words are sequentially arrayed, then phrase array becomes abc, bed, cde, def, efg. It is of course that the arrays including both two words and three words can be used.)

In the meantime, the phrase passing through the morpheme analyzer is names as a term. The term can be differentiated according to a setting value in the morpheme analyzer. At this time, an example of the setting value includes that 1) an uninflected word is only separated, 2) an inflected word is only separated but a citation form is only extracted, and 3)a citation form is only extracted in an adjective or adverb. Next, the phrase-based counting engine essentially generates (phrase, document number) or (phrase, counting value) information according to the generated phrase, preferably (phrase, counting value, document number) or (phrase, document number, location field name (e.g. the claim) in the document number) information, more preferably (phrase, counting value, document number, location field name (e.g. the claim) in the document) information. The phrase-based counting engine stores the generated information in the representing phrase DB 207-2. Table 5 represents an example of data included in the representing phrase DB 207-2. Table 5 represents the example of data generated by means of the specific word according to the nations. [Table 5]

The data represented in Table 5 are generated from patent documents 10-2003-0012345 and 10-2003-0056789, and it can be identified that the phrase generating engine generates the phrase in the unit of the 3 words. Further, it can be identified that the phrase of abc is present two times in the claims and five times in the detailed description of the invention. Further, it can be identified that the phrase of bed is present three times in the claims and four times in the detailed description of the invention. Further, it can be identified that the phrase of abc is present one time in the claims and two times in the detailed description of the invention of the application the patent document 10-2003-0056789. Therefore, the phrase-based counting engine of the present invention can generate the data of Table 5 for every repetitively or recursively obtained patent document. Further, the phrase generating engine can generate the word with the unit of two words, and it is obvious that the phrase-based counting engine of the present invention generates data of Table 11 for the phrase generated with the unit of two words. Further, the phrase-based counting engine generates data of Table 5 by means of the phrase generated in the unit of one word or four to five words.

At this time, the generation of an ID according to the phrase may be a problem. A method of generating the ID according to the phrase is well known in typical DB technology, but several exemplary methods will be suggested. First method is that an ID is sequentially granted for an initially generated phrase, a next generated phrase is searched in the existing phrase, and if the next generated phrase does not exist, a new ID is granted, and if the next generated phrase exists, the existing ID is granted. Such the processes are repetitively and recursively performed for every obtained phrase. The second method is that a temporary ID is sequentially granted for every obtained phrase without granting the ID for the phrase, the data of Table 11 is generated, and an identical formal ID is granted to an identical phrase while reading every phrase. Such the processes are repetitively and recursively performed for every phrase that has not been granted the formal ID.

Further, the phrase-based counting engine of the present invention can generate a phrase absolute ID whenever every word is generated and processed and grants the generated phrase absolute ID to every word.

Further, the phrase-based counting engine of the present invention can generate a word-based accumulated counting value on a basis of the target document set. It is obvious that in the target document set, the accumulated counting value can be calculated based on one document as represented in the above table, and also based on the specific field (e.g. the claim) within one document.

Further, the phrase-based counting engine of the present invention can generate a word-based accumulated counting value based on the entire document set and store it by a manner of Table 6 (In the above example, it is assumed that the entire document set includes patent documents 10-2003- 0012345 and 10-2003-0056789, and the numerical value is inputted based on the above table, and in actually, the inventive idea of the present invention is applied for the large entire document set.) [Table 6]

Further, the phrase-based counting engine of the present invention can generate the respective phrase-based accumulated counting values as represented in Table 7 with reference to the specific field (e.g. the claim) in the document.

Next, the representing phrase extracting engine of the present invention will be described. If there is the (phrase, document number) data extracted from the representing phrase extraction entire document set, it is easy to generate the (phrase, counting value, document number) data as indicated in Table 8 with the data. The (phrase, counting value, document number) data can be generated by increasing the counting value by 1 if an identical word is present. The representing phrase extracting engine of the present invention can generate (phrase, counting value, document number) data on a basis of (phrase, document number) data and (phrase, counting value, document number, location field name in the document number) data on a basis of (phrase, document number, location field name in the document number) data. Such the process can be performed for at least one representing phrase extraction entire document set that is a mother set of the representing phrase extraction target document set, as well as for every document included in at least one representing phrase extraction target document set.

The representing phrase extracting engine can generate (phrase, document number) data or (phrase, document number, location field name in the document number) data for the representing phrase extraction entire document set including 1) every document included in one nation DB, 2) every document corresponding to at least one given patent classification code, 3) every document of the specific applicant, 4) every document for the specific inventor, or 5) every document generated by limiting a term with respect to respective 1) to 5), or 6) every document generated by combining 1) to 5), respectively (including any set operation, such as union, difference component, intersection). In the meantime, the representing phrase extracting engine can generate (phrase, document number) data or (phrase, document number, location field name in the document number) data for every document corresponding to any subset of the representing phrase extraction entire document set.

At this time, the representing phrase extracting engine calculates frequency F in the preset representing phrase extraction target document set and 2) total frequency T in the representing phrase extraction entire document set that is a mother set of the representing phrase extraction target document set according to the phrase with respect to at least one preset representing phrase extraction target document set of at least one representing phrase extraction entire document set. It is of course that all frequency of phrase A can be calculated in the representing phrase extraction entire document set and all frequency of target set AT of the phrase can be calculated in the representing phrase extraction target document set.

At this time, the representing phrase extraction engine extracts the representing phrase of the representing phrase extraction target document set given in the preset representing phrase extraction entire document set referring to the representing phrase extraction policy DB. At this time, the representing phrase extraction policy is that whether the probability value of the phrase in (phrase, probability value of the phrase) is included in the preset reference range under the preset condition. An example of the probability value is F/T, F/A, and F/AT value according to the specific phrase. In the meantime, the representing phrase extraction policy is if 1) a variation value (increase/decrease rate, increase/decrease speed) of F, T, A, and AT value and 2) a variation value (increase/decrease rate, increase/decrease speed) of F/T, F/A, and F/AT value satisfies the preset reference range. At this time, the preset condition of the representing phrase extraction policy can correspond a value of another reference range to the probability value in (phrase, the probability value of the phrase) according to the term, nation, applicant, inventor, at least one patent classification code, or document set holding in common the preset attribute. For example, the value according to the different reference range can be applied to the probability value for the phrase extracted from the IPC H section including a lot of patent applications and the probability value of the words extracted in IPC D section including a small number of patent applications. A different reference can be flexibly applied to 1) the variation value (increase/decrease rate, increase/decrease speed) of F, T, A, AT value and 2) the variation value (increase/decrease rate, increase/decrease speed) of F/T, F/A, F/AT value according to the attribute of the document set (representing phrase extraction entire document set, representing phrase extraction target document set) from which the phrase is extracted.

Hereinafter, the concept of the representing phrase extraction of the present invention will be described with reference of a simple table. The representing phrase extracting engine of the present invention 1) corresponds the probability value according to the phrase ID (frequency of the specific phrase/the number of total phrase) on a basis of the frequency of the total phrase, selects the words corresponding to a preset probability value range as a candidate representing phrase, refers a preset representing phrase selection rule, and selects the representing phrase according to the target document set. It is obvious that the frequency can be calculated according to the specific field. In the example of Table 8, it is possible to count the words-based frequency according to the document number and extract the representing phrase corresponding to the preset representing phrase extraction policy according to the document number. [Table 8]

As shown in Table 8, there is an absolute phrase ID that one-to- one corresponds to every word. Therefore, the (phrase, document number) data essentially equals to the (phrase absolute ID, phrase, document number) data. The phrase ID corresponds to the respective absolute phrase IDs, the document number including the phrase corresponds to every absolute phrase ID, and the bibliographical particulars correspond to the respective document numbers. Therefore, the document number and the phrase ID-based frequency according to the document number are obtained with the document number in the target document set. In the same manner, the document number place can corresponds to the IPC of the preset level, the specific applicant, or the specific inventor, and the phrase ID-based frequency according to the target document set can be counted in each case. The counting of the phrase ID-based frequency according to the target document set can be implemented through a join command according to the table in the DBMS.

At this time, for example, when a selection rule of the representing phrase extraction policy included in the representing phrase extraction policy DB is that 1) the frequency (T) with respect to the total phrase numbers (e.g. 1,000,000,000) is less than 1/1000%, 2) the frequency in the target document number is equal to or over 1/1,000,000%, 3) the frequency T/total frequency TA ranges from 10% to 1%, it is examined if the respective phrases can be the representing phrase of the respective document numbers. The phrase of abc cannot be for any one of document numbers #1 to #4 (violates rule I)), the phrase of bed can be only for document number #2 (violates rule 2) for the document numbers 1 and 4), and the phrase of cde can be only for document number #4 (violates rule 3) for the document number #1). Therefore, the representing phrase of document number #1 does not include the phrase of abc, bed, and cde, the representing phrase of document number #3 does not include the phrase of abc, and the representing phrase of document number #4 includes the phrase of cde. According to the above method, the representing phrase extraction pre-processing module of the present invention can generate the words- based counting data by using the an indexer 401-3 of the search engine and the index and extract the representing phrase from the content according to the document, the preset document set, and the specific field, or all the specific fields extracted from the preset document set on a basis of the phrase-based counting data with the aforementioned method. For example, in the above paragraph, the phrase of bed can be corresponded to the representing phrase of document number #2. Next, if the representing phrase is extracted according to the representing phrase extraction target document set, the representing phrase can be stored in the unit of the representing phrase extraction target document set or in the unit of the individual document included in the representing phrase extraction target document set. That is, at least one representing phrase can be corresponded to one representing phrase extraction target document set unit or the individual document. At least one representing phrase corresponds in the unit of the individual field of the representing phrase extraction target document set or in the unit of the individual filed of the individual document included in the representing phrase extraction target document set. For example, the phrase of bed is corresponded to document number #2 to be stored. At this time, when document number #2 is called, the phrase of bed can be indicated.

At this time, the specific document number is assigned too many representing phrase. That is, there may be 100 representing phrases satisfying the above exemplary rule for the specific document, and in this case, 4) a rule of extracting 10 to 30 representing phrase according to a higher sequence of the frequency calculated for every representing phrase is applied so as to limit the number of the representing phrase. In the meantime, if there may be one or no representing phrase for the specific document, above rule 3) in the exemplary rules is relieved and applied to this case so as to maintain 10 to 30 numbers of the representing phrases. If the above procedure is repeatedly performed for every document or every documents set, the representing phrase within the preset number range (e.g. 10 to 30) can be corresponded to every document. Therefore, through the above procedures, it is possible to obtain the representing phrase corresponding to the document number. Further, the representing phrase and the calculation value for the representing phrase (e.g. the frequency in the document of the document number, the frequency in the total documents, etc.) are treated as an example of the bibliographical particular for the document number so as to be utilized for various analyses.

In the above embodiment, the frequency is calculated in the unit of the document, but the frequency also can be calculated in the unit of the content included in 1) the document set, 2) the field within the document (the claim, etc.), or 3) the field within every document included in the document set (e.g. claim 1 in the claims included in every document corresponding to IPC H04B 7/02 filed from the year from 2000 to 2005) in Korea. At this time, if the field is included in the inside of one document, it can be equally treated as one document (in view of the document number correspondence). However, in two document sets or more, it seems that there is no correspondence of the document number, but this problem can be easily solved by using three methods below: 1) the representing phrase is extracted in the unit of the individual document included in the document set, the union operation is performed for the representing phrase, and the number of representing phrase is limited in the preset range (e.g. from 10 to 30); 2) if the representing phrase has been already extracted according to the document number, the representing phrase corresponding to the document number included in the document set are obtained and the union operation is performed for the representing phrase, and the number of representing phrase is limited in the preset range; and 3) every document included in the document set is processed and then the representing phrase is extracted with corresponding to the unique ID of the document set, not to the document number. In the unit of the contents included in the field of the inside of every document of the 3) document set, 1) the summing-up process and 2) the document set unique ID corresponding method may be used, and if the representing phrase is given in the unit of the field, 3) the post-obtain summing-up method may be used.

Hereinafter, a method of extracting the representing phrase by using the indexer 401-3 of the search engine connected to the morpheme analysis engine will be described. The reason of using the search engine is that the search engine generally has a great performance of counting the number of the search word values so as to easily convert the data indexed by the indexer 401-3 into the data for DB. The indexer 401-3 stores the obtained term, a document number including the term, or a field name representing the term, case by case. For example, (term 1, #1) represents that term 1 is obtained from the document number #1, and (CL:term 1, #1) represents that the term 1 is obtained from claims (Claim, CL) of document number #1. The index of the search engine stores a great number of data, such as (term 1, #1) or (CL:term 1, #1), and if term 1 is inputted for the search word or term 1 is inputted with limiting the search field as the claim, document number #1 is outputted as the search result. Meanwhile, the search engine very rapidly calculates how many term 1 is included in the entire documents or in the claims (typically, the search engine first returns the total numbers of the search results, and then shows a part of a front of the total number). At this time, the phrase generating engine of the present invention receives the n number of phrase/terms, which are passing the morpheme analysis engine (analyzer), combines the received n number of terms by a preset manner, transmits the combined terms to the indexer 401- 3 of the search engine together with the document number (at this time, if the indexer has already obtained the document number, it is not necessary to transmit the document number. For convenience of understanding, the document number is included and transmitted as the source of the transmitted combined term). At this time, a method of combining the terms includes any one of 1) combining, in the unit of one sentence, all the terms obtained from a single sentence on a basis of 2 to 5 terms with any available method, 2) combining all the terms included the entire document- baseds on a basis of 2 to 5 terms with any available method (in this case, the combination of the terms is very large so that it takes much time for processing, but it is possible to generate the most fine word set.), and 3) sequentially combining the terms on a basis of 2 to 5 terms in a single sentence. Of course, the method of processing in the unit of a paragraph of the specific field (the claim, etc.) can be identically treated with the method of processing the entire document-based.

For example, the phrase generating engine of the present invention obtains the outputted phrase and generates the phrase array including 1 to 5 words (preferably, 2 to 3 words). For example, if one sentence included in a single patent document is inputted, the outputted words are assumed as abcde f, g in sequence according to the morpheme. At this time, if the phrase generating unit of the phrase generating engine includes 3 words, the phrase generating engine sequentially generates the phrase, abc, bed, cde, def, and efg according to the word, or phrase including two words, such as ab, ac, ad, ae, af, ag, be, ..., and fg, which is previously described. At this time, the phrase generating engine transmits the phrase (multiple terms), such as abc, bed, or ab, ac to the indexer 401-3. The indexer 401-3 stores the transmitted phrase (multiple terms) in the form of (abc, #1), (bed, #1), (ab, #1), (ac, #1), etc. At this time, if the fields are limited, the indexer 401-3 includes even the specific field name presenting the phrase (multiple terms), such as (CL : abc, #1), (CL : bed, #1), (CL : ab, #1), (CL : ac, #1) so as to store them in the index. Table 9 below is a conceptual table illustrating an example of the index. [Table 9]

In Table 9, D represents the detailed description of the invention and C represents the claims.

In table 9, the word of abc is present three times in the detailed description and two times in the claim of document number #1, the phrase of bed is present one time in the detailed description and one time in the claim of document number #1, and the phrase of ac is present one time in the detailed description of document number #1. In the meantime, the phrase of abc is present two times in the detailed description and one time in the claim of document number #2. At this time, the phrase-based counting engine of the present invention receives the data of the index and generates the data relating to the number of phrases (multiple terms) and the document number serving as the source of the phrase. The generated data is identical or corresponding to the data generated by the aforementioned method, and the example for the index is shown in Table 10.

[Table 10]

Through the above method, the representing phrase extraction preprocessing module of the present invention generates the phrase-based counting data by utilizing the indexer 401-3 of the search engine and the index and extracts the representing phrase from the contents according to the documents, the preset document sets, the specific field, or every specific field extracted from the preset document set.

Next, if the representing phrase is extracted according to the representing phrase extraction target document sets, the representing phrase can be stored in the unit of the representing phrase extraction target document set or in the unit of the individual document included in the representing phrase extraction target document set. That is, at least one representing phrase can correspond to one representing phrase extraction target document set or the individual document. At least one representing phrase can correspond to the unit of the individual field of the representing phrase extraction target document set or the unit of the individual field of the individual document included in the representing phrase extraction target document set.

*498It has been described that the representing phrase extracting engine of the present invention also generates the document number information corresponding to the phrase. This is for the purpose that, if the document number is corresponding to the phrase, it is easy to extract the document set in which the phrase serves as the representing phrase. If the document set in which the phrase serves as the representing phrase can be easily extracted, the document set is determined as the to-be-analyzed document set, to be utilized for various analysis. For example, for one document set in which the phrase of "radio frequency identification tag" is extracted as the representing phrase, one document set, various analysis supported by the analysis module and the patent intelligence module 60 of the present invention according to 1) the various analysis indexes, such as the number of applications/the number of registrations/the rate of registration/occupancy rate/concentration rate/according to the nation/year/applicant, 2) distribution analysis according to the multi-stages (depth) of the IPC or other patent classification code according to the nations/years/applicants, and 3) citing and cited analysis according to the detailed document set (according to the applicant, patent classification code) of the document set, can be performed. Further, if the document number does not correspond to the word and the document including the specific phrase is extracted through the search engine, it may occur a problem of failing to discriminate if the specific phrases are the representing phrase of the document .

The representing phrase information pre-processing method is well illustrated in FIGs. 105 to 107. Referring to FIG. 105, a method of generating the combination for extracting the representing phrase through the representing phrase extraction pre-processing module will be briefly described.

The representing phrase extraction pre-processing module obtains the document set including at least two patent documents (S5520), processes the content according to the entire individual patent documents or various fields extracted from the document set with the morpheme analyzer so as to extract at least two phrases (term or terminology) (S5530), generates all the combinations of the phrase with the preset method based on the preset reference (S5540), and stores the generated combination of the phrase in the DB or transmits it to the search indexer 401-3 together with the patent document information (S5550), so as to generate the combination of the phrase for extracting the representing phrase.

A method of generating the combination of the phrase for extracting the representing phrase according to the field / the partial document set through the representing phrase extraction pre-processing module and extracting the combination satisfying the preset condition will be briefly described with reference of FIG. 106.

The representing phrase extraction pre-processing module obtains the document set including at least two patent document (S5620), processes the content according to the entire individual patent documents or various fields extracted from the document set with the morpheme analyzer so as to extract at least two phrases (term or terminology) (S5630), generates all the combinations of the phrase with the preset method based on the preset reference (S5640), and stores the generated combination in the DB or transmits it to the search indexer 401-3 together with the patent document information (S5650), obtains the bibliographical particular of the respective documents included in the document set and generates the document sub-set based on at least one reference (S5660), and obtains the number of phrase combinations limited to the document subset and the number of phrase combinations included in the document (entire) set so as to extract the phrase combination satisfying the preset reference according to the entire document or various fields (S5670).

A method of generating the combination for extracting the representing phrase through the representing phrase extraction pre- processing module and comparing the generated combination according to the respective document sets will be briefly described with reference to FIG. 107.

The representing phrase extraction pre-processing module obtains multiple document sets including at least two patent documents (S5720), processes the contents according to the entire individual patent documents or various fields extracted from the document set by the morpheme analyzer according to the respective document sets so as to extract at least two phrase (term) (S5730), generates the combination of the extracted phrase (S5740), and stores the generated combination in the DB or transmits it to the search indexer 401-3 together with the patent document information

(S5750), and compares the phrase combination according to the multiple document sets based on the preset reference (S5760).

Multi-patent classification code correlation pre-processing module

There are many patent documents, each of which is given at least two patent classification codes belonging to one patent classification code system. This trend, that is, the trend of giving multiple patent classification codes to one patent document gets stronger according to: 1) enlargement of the range for the patent application; 2) advancement in the level of the invention; 3) trend of technology convergence; and 4) accuracy increase in granting the patent classification code. The fact that a single patent document is granted at least two patent classification codes may imply that the corresponding patent document includes at least two technical views, at least two technical constructions, at least two technical contents, at least two technical aspects, at least two technical attributes, and at least two technical characteristics corresponding to said at least two patent classification codes. Therefore, it is necessary to treat a document granted at least two patent classification codes in a way different from that for a document granted a single patent classification code.

The multi-patent classification code correlation pre-processing module of the present invention is intended to find and use the correlation hidden between the multiple patent classification codes, which have been granted to a single patent document. The multi-patent classification code correlation pre-processing module includes a homogeneous multi-patent classification code correlation pre-processing module and a heterogeneous multi-patent classification code correlation pre-processing module. The homogeneous multi-patent classification code correlation pre-processing module is intended to find the correlation hidden between homogeneous patent classification codes given to a single patent document based on one of the IPC, UPC, and FT (i.e. based on the same patent classification code system), and the heterogeneous multi- patent classification code correlation pre-processing module is intended to find the correlation hidden between multiple patent classification codes given to a single patent document based on at least two patent classification code systems selected from the IPC, UPC, and FT. FIG. 60 shows the configuration of the multi-patent classification code correlation pre-processing module, the configuration of the homogeneous multi-patent classification code correlation pre-processing module, and the configuration of the heterogeneous multi-patent classification code correlation pre-processing module.

Homogeneous multi-patent classification code correlation pre- processing module

Hereinafter, the homogeneous multi-patent classification code correlation pre-processing module of the present invention will be described. Even when a single document is granted multiple patent classification codes of the same type or two or more types of patent classification codes, multiple patent classification codes on a selected single patent classification code system are processed by the homogeneous multi-patent classification code correlation pre-processing module. The multiple patent classification codes are processed by the homogeneous multi-patent classification code correlation pre-processing engine of the present invention, and the result of the processing is stored in a homogeneous representing phrase multiple patent classification code correlation DB 207-3 of the present invention.

A theoretical description will be given first, and a detailed example will be then discussed. First, a simplest model when a single document includes two patent classification codes will be discussed in order to present a method for processing the correlation between multiple patent classification codes included in a single document. Let us assume that a single patent document is given patent classification code Al and patent classification code Bl. Also, in the patent classification code system, the direct higher node (parent patent classification code) of Al is A2 and the direct higher node of A2 is A3. According to this rule, higher nodes of Al may include A2, A3, A4, ..., and An. In the same manner, Blmay be allocated B2, B3, B4, ... , and Bn as its higher nodes. From among these Ans or Bns, the uppermost one serves as a section. Further, let us assume that a patent classification code system name, such as IPC, exists above the section. Then, the Ans and the Bns have at least one common node under at least a section or a patent classification code system named IPC (even when they are very far from each other, for example, when their sections are different from each other, they have at least one common node because they share the same patent classification code system named IPC). From among the common nodes, a node of the lowermost level in the patent classification code system is named a lowermost common patent classification code. At this time, if the lowermost common patent classification codes of Al and Bl are (A(i+1)) and (B(j+1)), respectively, then A(i+1) = B(j+1). In view of the contents of the given single patent document, when Aland Bl exist together, Al and Bl have a correlation with respect to the patent document (that is, an invention or inventions within a single patent document independently or commonly have the attributes of Al and Bl). At this time, due to the basic principle of the patent classification code system, Al has correlations with all the parent nodes of Bl, which include B2, B3, ... and B(j+1). In the same way, Bl has correlations with all the parent nodes of Al, which include A2, A3, ..., and A(i+1). Here, since B(j+1) is equal to A(i+1), which itself is a parent node, it is unnecessary to discuss the correlation between Al and B(j+1). Also, it is unnecessary to discuss the correlations between all As and B(j+l)(=A(i+D), because the latters inevitably exist in parent nodes. That is to say, it is preferable to omit consideration of a lowermost common parent node in a correlation calculation table.

When the homogeneous multi-patent classification code correlation pre-processing module processes the multiple patent document codes, it can generate the data as shown in Table 11 by referring to the patent classification code system data (tree structure). Here, although A(i+1) and B(j+1) are unnecessary cells in an actual table (only the cells up to Ai and up to Bj are necessary in the correlation table), they also are indicated in the table for convenience of description (in order to show that they are omitted in the cells). That is, an actual table includes only the cells up to Ai and Bj .

[Table 11]

That is, it can be said that all the intersection couples (for example, (Al, Bl), ... (Ai, Bj), etc.) have correlations through the given patent document.

Korean Patent Application No. 10-2005-0111868 dated January of 2006 is granted H04B 7/26 and H04B 7/15. The following description is based on this example. Parents of H04B 7/26 include, in turn, H04B 7/24, H04B 7/00, H04B, H04, and H. Parents of H04B 7/15 include, in turn, H04B 7/14 and H04B 7/00. Further, the lowermost common patent classification code is H04B 7/00. Therefore, these codes can be arranged as shown in Table 12 below. As described above, cells relating to H04B 7/00 have no values.

[Table 12]

Therefore, based on Korean Patent Application No. 10-2005-0111868, nine patent classification code correlation intersection couples are generated. Here, if one document has at least two patent classification codes, the frontmost patent classification code is called a main patent classification code. Therefore, in the present document, H04B 726 is the main patent classification code, other higher patent classification codes at parent nodes of the main patent classification code are also treated as main patent classification codes in the given patent document. The main patent classification code plays an important role in treating a document having at least three patent classification codes. In the case of a document having at least three patent classification codes, the patent classification codes can be treated in the ways as follows.

First, all the patent classification codes can be treated equally. When a document has n patent classification codes, there are nC2 methods of taking a combination of two patent classification codes from the n patent classification codes according to the mathematical combination theory. For each of the nC2 patent classification code combinations, the treatment corresponding to the two patent classification codes is performed, so as to obtain nC2 tables, each of which has information on all intersection couples.

According to the next method, a main patent classification code is first determined and taken, and the other patent classification code is then taken from (n-1) sub-patent classification codes. Then, (n-1) combinations and (n-1) tables are obtained, and all patent classification codes in each table are treated equally. When a document has n patent classification codes, there are nC2 methods of taking a combination of two patent classification codes from the n patent classification codes according to the mathematical combination theory. For each of the nC2 patent classification code combinations, the treatment corresponding to the two patent classification codes is performed, so as to obtain nC2 tables, each of which has information on all intersection couples. At this time, an intersection couple relating to the main patent classification code may be assigned a weight relating to the main patent classification code. Further, with respect to (n-1) sub-patent classification codes, (n-l)C2 combinations are obtained. Then, each combination is subjected to the treatment as described above, so as to obtain information on all the intersection couples. Here, an intersection couple relating to a sub- patent classification code may be assigned a weight relating to the sub- patent classification code.

Korean Patent Application No. 10-2006-0012606 has patent classification codes assigned on January of 2006, which include H04B 7/04, H04B 7/155, and H04Q 7/30. Parent nodes of H04B 7/04include H04B 7/02, H04B 7/00, H04B, H04, and H, parent nodes of H04B 7/155 include H04B 7/15, H04B 7/14, H04B 7/00, H04B, H04, and H, and parent nodes of H04Q 7/30 include H04Q 7/20, H04Q 7/00, H04Q, H04, and H. Here, the lowermost common patent classification code for a combination of H04B 7/04 and H04B 7/155 is H04B 7/00, the lowermost common patent classification code for a combination of H04B 7/155and H04Q 7/30 is H04, and the lowermost common patent classification code for a combination of H04B 7/04 and H04Q 7/30 is H04. Here, if all the patent classification codes are equally treated, X2 tables as follows are obtained. First, Table 13 below is a correlation table of H04B 7/04 and H04B 7/155.

[Table 13]

Next, Table 14 below is a correlation table of H04B 7/04 and H04Q

7/30.

[Table 14]

Next, Table 15 below is a correlation table of H04B 7/155 and H04Q

7/30. [Table 15]

Hereinafter, methods of giving a weight are described. In large, it is possible to give a weight in two ways. First, as described above, a larger weight may be given to a correlation table relating to the main patent classification code while a smaller weight is given to a correlation table relating to the sub-patent classification code. For example, a correlation table relating to the main patent classification code may be given 1 while a correlation table relating to the sub-patent classification code is given 0.5.

*Then, the correlation table of H04B 7/155 and H04Q 7/30 is expressed by Table 16 below.

[Table 16]

Second, a uniform weight may be used for a patent document having multiple patent classification codes. For example, let us assume that, when a patent document has two patent classification codes, each patent classification code has a correlation couple value of 1. Then, in the case of a patent document having n patent classification codes, it is possible to uniformly allocate a value obtained by dividing the correlation couple value by nC2, which is the number of possible combinations, to the patent classification code couples. That is, it is possible to uniformly allocate a weight of l/(nC2). Then, the correlation table of H04B 7/155 and H04Q 7/30 is expressed by Table 17 below.

[Table 17]

Third, for a document having at least three patent classification codes, different weights may be assigned to a couple relating to the main patent classification code and a couple relating to the sub-patent classification code. For example, when there are n patent classification codes, there are (n-1) combinations relating to the main patent classification code. Therefore, it is possible to allocate 75% of the weight to combinations relating to the main patent classification codes while allocating a weight of 0.75*1/(n-1) to relevant couples relating to them. When 25% of the weight is given to all combinations relating to the sub-patent classification codes, it is possible to allocate a weight of 0.25*l/((n-l)C2) to all couples relating to the sub-patent classification codes. At this time, distribution of the percentage (%) may be controlled by using preset values. For example, it is possible to allocate 50% to couples relating to the main patent classification codes and 50% to couples relating to the sub-patent classification codes when n > 5, to allocate 60% and 40% to couples relating to the main patent classification codes and the sub-patent classification codes when n= 4, and to allocate 75% and 25%to couples relating to the main patent classification codes and the sub-patent classification codes when n=3. That is, when the number of IPC codes is larger than 3, there may be a (sub IPC, Sub IPC) patent classification code combination couple. Therefore, it is possible to give different weights to a (main IPC, Sub IPC) patent classification code combination couple and a (Sub IPC, Sub IPC) patent classification code combination couple, respectively. By allocating 75% and 25% to couples relating to the main patent classification codes and the sub-patent classification codes when n=3, it is possible to generate the Tables as follows. First, Table 18 below is a correlation table of H04B 7/04 and H04B 7/155.

[Table 18]

Next, Table 19 is a correlation table of H04B 7/04 and H04Q 7/30.

[Table 19]

Next, Table 20 is a correlation table of H04B 7/155 and H04Q 7/30.

[Table 20]

Meanwhile, based on an assumption that documents each having multiple patent classification codes have the same contribution degree (e.g. 1), it is possible to take the weight granting method as follows into consideration. As an extreme example, Korean patent application No. 10-2005-0042032 has two patent classification codes including H04B 7/02 and H04B 7/14, which are in a next sibling relation (that is, H04B 7/02 and H04B 7/14 have the same parent node, i.e. H04B 7/00, which is the direct higher patent classification code). In the case of H04B 7/02 and H04B 7/14, there is only one correlation table, and the table has only one cell. Table 21 below corresponds to the correlation table for the Korean patent application No. 10-2005-0042032.

[Table 21]

It can be said that H04B 7/02 and H04B 7/14 are strongly related to each other for the particular patent document. In the meantime, the Korean Patent Application No. 10-2006-0012606 has three tables, which have 6 cells, 16 cells, and 20 cells, respectively. Here, based on the single patent classification code combination, i.e. (H04B 7/02, H04B 7/14), an important issue is how to appraise the contribution degree of the Korean patent application No. 10-2005-0042032 and the contribution degree of the Korean Patent Application No. 10-2006-0012606. The weight of the contribution degree may be treated as follows.

First, in any document regardless of the table or the number of cells within the cell based on existence or occurrence of the correlation, a combination (Ai, Bj), such as (H04B 7/02, H04B 7/14) described above, of particular patent classification codes can be treated with the same contribution degree or weight (e.g. 1).

Second, in consideration of the number of tables and the number of cells, different contribution degrees or weights may be granted to particular patent classification code combinations (Ai, Bj) according to the patent documents. Here, the weight or contribution degree to the particular patent classification code combination (Ai, Bj)may have a value obtained by dividing 1 by the number of all cells of all tables of each patent document. For example, in relation to the combination (H04B 7/02, H04B 7/14), the weight or contribution degree relating to Korean patent application No. 10-2005-0042032 is 1 (which implies that there is one table and one cell), and the weight or contribution degree relating to Korean Patent Application No. 10-2006-0012606 is 1/(6+16+20, which corresponds to the number of all cells in the three tables).

Third, the weight indicating the contribution degree may have a value obtained by dividing 1 by the number of all cells within the tables relating to the particular patent classification code combination (Ai, Bj). For example, in relation to (H04B 7/02, H04B 7/14), the weight or contribution degree relating to the document of the Korean patent application No. 10-2005-0042032 has a value of 1 (which implies that there is one table and one cell), and the weight or contribution degree relating to Korean Patent Application No. 10-2006-0012606 is 1/(6, which corresponds to the number of all cells in the three tables).

Fourth, there is a method of calculating the weight or contribution degree inversely to the second or third method. That is, the weight or contribution degree may have a value obtained by multiplying 1 by the number of all cells within the tables relating to the particular patent classification code combination (Ai, Bj). For example, in relation to (H04B 7/02, H04B 7/14), the weight or contribution degree relating to the document of the Korean patent application No. 10-2005-0042032 has a value of 1 (which implies that there is one table and one cell), and the weight or contribution degree relating to Korean Patent Application No. 10-2006-0012606 is 6, which corresponds to the number of all cells in the three tables). Such a weight granting method is expected to be useful for calculation of convergence degree of heterogeneous technologies. That is, when a table includes a small number of cells, concerned technologies may correspond to strongly-related technologies bound together by the technology classification system itself. Therefore, the larger a table, the larger the distance between the two technologies in the technology classification system.

By the methods described above, it is possible to acquire the following information in relation to the patent classification code combination (Ai, Bj).

First, counting information in relation to (H04B 7/02, H04B 7/14) (Ai, Bj) is discussed. By granting a weight mentioned in the present invention or matching the concept of the present invention during the counting, it is possible to acquire weight-based counting information in which at least one weight is applied to the patent classification code combination (Ai , Bj ).

Second, every patent classification code combination contains information on a document, based on which the combination is created. That is, the patent classification code combination (H04B 7/02, H04B 7/14) corresponds to a combination created based on Korean Patent Application No. 10-2006-0012606 and Korean patent application No. 10-2005-0042032, and the patent classification code combination (H04B 7/15, H04B 7/24) corresponds to a combination created based on Korean Patent Application No. 10-2005- 0111868. Therefore, a patent classification code combination (Ai, BjCorresponds to a specific document number, which may correspond to all information in relation to the document including all bibliographical information of the document having the specific document number. Every table as described above has two relevant patent classification code combinations and corresponds to a patent document having the patent classification code combinations, and the patent document may correspond to bibliographical information of the patent document. That is, all patent classification code combinations (Ai, Bj) existing in the correlation table of H04B 7/155 and H04Q 7/30 correspond to Korean Patent Application No. 10-2005-0111868 as a specific document number. Further, by way of an example, in bibliographical information of the document corresponding to the specific document number, an application number of the document is 10-2006-0012606, an applicant thereof is Samsung electronics, inventors thereof are Do-in Choi and Seong-Tak Hwang, and application date thereof is February 9, 2006. As in this example, the document may correspond to all information of the document, which includes bibliographical information and a specific document number. Of course, the specific document number may be an application number. Also, if the present patent application has been filed in the United States too, such information described above may be mapped to or correspond to the bibliographical information of the US application.

It is preferred that the above-described corresponding relation between all the patent classification code combinations (Ai, Bj)and the document number is stored together with the counting value in a database during the creation of the combinations.

Ill When the obtained patent document has at least two patent classification codes, the homogeneous multi-patent classification code correlation pre-processing module creates a predetermined number of tables for each pre-patent classification code combination by referring to the patent classification code system data (tree structure) of each patent classification code combination. Here, numeric values in each table may be values reflecting weights based on a predetermined standard. Then, the homogeneous multi-patent classification code correlation pre-processing module may create patent classification code combination (Ai, Bj)-based correlation result data as shown in Table 22 below by applying a predetermined weight to the patent classification code combinations (Ai, Bj) extracted from each table.

In order to discuss the present model, let us assume that there are only three patent documents including Korean Patent Application No. 10-2005-0111868, Korean Patent Application No. 10-2006-0012606, and Korean patent application No. 10-2005-0042032 (of course, by repeatedly processing all the obtained documents, it is possible to generate the following data in relation to all the patent classification code combinations (Ai, Bj) based on all the documents). Here, for the patent classification code combination (Ai, Bj), it is expected to be possible to create the table as follows.

[Table 22]

In Table 22, n denotes the number of patent classification codes),

(M,S) refers to (main IPC, Sub IPC), (S₁S) and refers to (sub IPC, Sub

IPC). When n>2, differential allocation was applied to (M,S) and (S₁S). Specifically, (M,S) was allocated a weight of 75%/(n~l), and (S,S) was allocated a weight of 25%/(n-l)C2. When n=2, simply 1 was granted.

*In view of the patent classification code combination (H04B 7/14,H04B 7/02), it is noted that two document numbers correspond to the combination. That is, it implies that the patent classification code combination couple (H04B 7/14.H04B 7/02) has simultaneously occurred in the two documents, and its frequency (simple frequency) has a value of 2. It is possible to use predetermined methods of granting a weight or contribution degree, such as l/nC2 uniform allocation and (M,S) and (S,S) differential allocation.

Table 22 contains correlation data created for the patent classification code combination (Ai, Bj) in relation to three patent documents including Korean Patent Application No. 10-2005-0111868, Korean Patent Application No. 10-2006-0012606, and Korean patent application No. 10-2005-0042032. It is obvious to one skilled in the art that it is possible to repeatedly or recursively apply the above method to all the obtained document. It is also obvious to one skilledin the art that it is possible to create correlation data for each patent classification code combination (Ai, Bj) based on at least one of preset weights or contribution degrees.

As described above, the homogeneous multi-patent classification code correlation pre-processing module may, for a couple (Ai, Bj), store at least of 1) a document number containing the (Ai, Bj) value, 2) weight (Wij) according to particular situation(i)-based weight policy(j), 3) whether (M,S) or (S,S), 4) other attribute information of (Ai, Bj) in a database.

It is natural that the method of processing multiple patent classification codes by the homogeneous multi-patent classification code correlation pre-processing module is not limited to the IPC. That is, in the completely same way as in the IPC, multiple patent classification codes of US patent documents may be processed by means of USPC, multiple patent classification codes of Japanese patent documents may be processed by means of FT and FI, and multiple patent classification codes of European patent documents may be processed by means of ECLA.

Technology convergence pattern analysis module

By processing at least one of 1) the document number of the document containing the value (Ai, Bj), 2) the weight (Wij) following the weight policy j based on the particular situation i, 3) whether (M,S) or (S,S), and 4) other attribute information on (Ai, Bj), which are stored by the homogeneous multi-patent classification code correlation pre^¬ processing module for the patent classification code combination (Ai, Bj), it is possible to calculate, obtain, or extract the information as follows. The following information is processed by the technology convergence pattern analysis module of the present invention. Essentially, the technology convergence pattern analysis module can process two types of information as follows.

First, the homogeneous multi-patent classification code correlation pre-processing module can obtain the number of times by which each patent classification code couple occurs. For example, (H04B 7/26, H04B 7/15) occurs only once in Korean Patent Application No. 10-2005- 0111868, while (H04B 7/14,H04B 7/02) occurs in both Korean Patent Application No. 10-2006-0012606 and Korean Patent Application No. 10-2005- 0042032 (when it is not a main patent classification code or when all patent classification codes are equally treated, (Ai, Bj) = (Bj, Ai). When one of the codes is a main patent classification code, (Ai, Bj) and (Bj, Ai) may be treated either equally or differently). Therefore, the homogeneous mult i-patent classification code correlation pre-processing module may have a record on the number of times of occurrence for all patent classification code couples. Further, the module may store patent document numbers of patent documents in which those couples occur.

Second, when the particular patent classification code is given, the homogeneous multi-patent classification code correlation pre^¬ processing module is expected to be capable of finding a patent classification code most frequently coupling with the particular patent classification code. This can be easily obtained by finding all the couples containing the given particular patent classification code and counting the number of times of occurrence of the couples. Here, by applying a preset weight or contribution degree weight, it is possible to find a patent classification code most frequently coupling with the particular patent classification code. Also, by referring to the patent classification code system (tree structure), it is possible to find a patent classification code most frequently coupling in a preset level. For example, the homogeneous multi-patent classification code correlation pre-processing module can find a patent classification code most frequently coupling with H04B 7/26, and can a patent classification code most frequently coupling with H04B 7/26 in the level of the IFC sub class, IPC main group, or IPC subgroup.

Third, based on the corresponding relation between the patent classification code couple and the patent document number, the homogeneous multi-patent classification code correlation pre-processing module can generate various analysis results by using the relation between the patent classification code couple and bibliographical information corresponding to the patent document number. The analysis results may be obtained through analysis of a combination between the patent classification code couple and at least one selected from 1) nation, 2) period, 3) applicant, 4) inventor, and 5) agent. For example, it is possible to obtain a most- frequent Iy occurring patent classification code couple among Korean Patent Applications of years from 2000 to 2005 of Samsung Electronics Inc. Further, when a patent classification code is given, the homogeneous multi-patent classification code correlation pre-processing module can find a patent classification code, which is most frequently combined with the given patent classification code. Also, when a patent classification code is given, the homogeneous multi-patent classification code correlation pre-processing module can find a patent classification code, which is most frequently combined with the given patent classification code, from a document set based on at least one selected from 1) nation, 2) period, 3) applicant, 4) inventor, and 5) agent. For example, the homogeneous multi-patent classification code correlation pre-processing module can find a patent classification code most-frequentIy combined with H04B 7/26 according to each stage in a patent classification code system, and can find a patent classification code most-frequentIy combined with H04B 7/26 in Korean Patent Applications of years from 2000 to 2005 of Samsung Electronics, Co. Ltd. By the analysis as described above, it is possible to understand a degree of technology convergence in patent specifications of a particular subject during a particular period.

Fourth, for a given patent classification code combination (Ai, Bj), the homogeneous multi-patent classification code correlation pre^¬ processing module can cause a particular patent document set to correspond to the given patent classification code combination (Ai, Bj). Then, the homogeneous multi-patent classification code correlation pre-processing module can perform various analyses for the corresponding patent document set, which yield results of analyses for the patent classification code set (Ai, Bj). By way of example, the particular patent document set may be: 1) a patent document set generated by specifying the patent document set by a particular classification code according to a particular patent classification code system in a database of a particular nation, 2) a patent document set generated by specifying the patent document set by a particular applicant name in a database of a particular nation, 3) a patent document set generated by specifying the patent document set by a particular inventor name included in a patent document of a particular applicant name (i.e. a particular applicant name and a particular inventor name) in a database of a particular nation, 4) a patent document set generated by specifying the patent document set by a particular agent name in a database of a particular nation, 5) a patent document set generated by specifying the patent document set by a particular applicant name and a particular classification code according to a particular patent classification code system in a database of a particular nation, 6) a patent document set generated by specifying the patent document set by a particular applicant name, a particular inventor name, and a particular classification code according to a particular patent classification code system in a database of a particular nation, 7) a patent document set generated by specifying the patent document set by a particular applicant name and a particular agent name in a database of a particular nation, 8) all patent document sets of a particular nation, 9) all patent document sets of at least two nations, 10) a patent document set specified by a predetermined period further to the conditions of 1) to 9), or 11) a patent document set specified by a particular condition, such as whether registered or not and whether a request for examination has been filed or not, further to the conditions of 1) to 9), and the patent document sets enumerated in 1) to 11) may be calculated and stored in advance. Of course, the pre-calculation and pre-storage may be performed for a document set specified or generated by a user.

Further, the homogeneous multi-patent classification code correlation pre-processing module may extract documents having at least two patent classification codes from at least one document set specified by the conditions of 1) to lDdescribed above, and obtain the patent classification code combination (Ai, Bj) from the extracted documents. For the obtained patent classification code combination (Ai, Bj), the homogeneous multi-patent classification code correlation pre-processing module generates correlation data as in the tables, and performs arrangement or analysis of the generated correlation data (e.g. analysis of frequency, a most-rare value, a maximum value, a minimum value, a number according to each year, and number according to each subject such as an applicant or an inventor). By the analysis as described above, it is possible to obtain information on technology convergence in view of the patent classification code combination (Ai, Bj) according to at least one document set specified by the conditions 1) to 11) or according to the extracted document set.

The above-described repeating execution method by the homogeneous multi-patent classification code correlation pre-processing module is only one example, and one skilled in the art can try various calculation methods according to spirit of the present invention. It is natural that the results (for example, tables having values for each patent classification code couple regardless of a granted weight) of the above process for all patent classification codes of all patent documents each having at least two patent classification codes may be either stored in a database or provided through a view realized in real time or a materialized view.

An example of the above-described process is shown in FIG. 103, which will be described hereinafter. The homogeneous multi-patent classification code correlation pre-processing module obtains a document set including at least one patent document (S5320), extracts multi-code documents having at least two patent classification codes from the obtained document set (S5330), generates multiple patent classification code combinations included in the extracted multi-code documents according to a predetermined standard for discriminated processing of the main patent classification code and the sub-patent classification code (S5340), obtains information on parent nodes of each patent classification code included in each of the patent classification code combinations in the patent classification code system (S5350), obtains a lowest common patent classification code for the parent nodes (S5360), generates parent node patent classification code couples including itself up to just before the lowest common patent classification code (S5370), stores each couple after granting the couple a uniform weight or a weight according to a predetermined standard (S5380), and generates a predetermined standard- based statistic value, parameter or calculation value for the stored couple (S5390).

Hereinafter, a method of pre-processing and then comparing multiple patent classification codes for at least two document sets by the homogeneous multi-patent classification code correlation pre-processing module will be described with reference to FIG. 104.

The homogeneous multi-patent classification code correlation pre^¬ processing module obtains at least two document sets each including at least one patent document (S5410), extracts a multi-code patent document having at least two patent classification codes from each document set (S5430), generates multiple patent classification code combinations included in the multi-code patent documents according to a predetermined standard for discriminated processing of the main patent classification code and the sub-patent classification code (S5440), obtains information on parent nodes of each patent classification code included in each of the patent classification code combinations in the patent classification code system (S5450), obtains a lowest common patent classification code for the parent nodes (S5460), generates parent node patent classification code couples including itself up to just before the lowest common patent classification code (S5470), stores each couple after granting the couple a uniform weight or a weight according to a predetermined standard (S5480), generates a predetermined standard-based statistic value, parameter or calculation value for the stored couple (S5490), and compares the statistic value, parameter or calculation value for each of the document sets (S5495).

FIGs. 115 to 117show examples of a method for analysis of multiple patent classification codes.

FIG. 115 illustrates an example of a method for presenting analysis information on most-frequent Iy converging patent classification codes based on an obtained patent classification code.

The homogeneous multi-patent classification code correlation pre^¬ processing module obtains at least one patent classification code (S6520), obtains information on most-frequentIy converging patent classification codes based on the obtained patent classification code (S6530), and displays the most-frequent Iy converging patent classification codes in at least one stage reflecting a patent classification code system or performs at least one analysis preset for a patent document set corresponding to the most-frequentIy converging patent classification codes (S6540).

FIG. 116 illustrates an example of a method for extracting most- frequent patent classification codes from an obtained document set, and presenting analysis information on most-frequentIy converging patent classification codes based on the extracted patent classification codes.

The homogeneous multi-patent classification code correlation pre^¬ processing module obtains at least one document set (S6620), obtains patent classification codes for each document of the document set (S6630), ranks the obtained patent classification codes for each document based on a most-frequent value in the level of the document set (S6640), obtains information on most-frequent Iy converging patent classification codes for each of the ranked patent classification codes (S6650), and displays the most-frequent Iy converging patent classification codes in at least one stage reflecting a patent classification code system or performs at least one analysis preset for a patent document set corresponding to the most- frequent Iy converging patent classification codes (S6660).

FIG. 117 illustrates an example of a method for extracting most- frequently converging patent classification codes from patent classification codes extracted from each document in an obtained document set, and presenting analysis information generated by synthesizing the extracted most-frequentIy converging patent classification codes.

The homogeneous multi-patent classification code correlation pre- processing module obtains at least one document set (S6720), obtains patent classification codes for each document of the document set (S6730), obtains information on most-frequentIy converging patent classification codes from the obtained patent classification codes for each document (S6740), synthesizes the obtained information on most-frequent Iy converging patent classification codes for each document (S6750), and displays the most-frequent Iy converging patent classification codes in at least one stage reflecting a patent classification code system or performs at least one analysis preset for a patent document set corresponding to the most-frequent Iy converging patent classification codes (S6760). The homogeneous multi-patent classification code correlation preprocessing module may be applied to simple patent classification code couples (PCl, ..., PCi, PCj,..., PCn) included in each document, without the inventive sprit including lower patent classification codes. In other words, without considering higher patent classification codes of a given PCi in a given patent classification code system, it is possible to configure a patent classification code combination or couple for only (PCi, PCj)s observed in documents. At this time, since PCl is a main patent classification code, it is possible to apply the above-described weight granting method between main codes and sub-codes as it is. Heterogeneous multi-patent classification code correlation preprocessing module

The homogeneous multi-patent classification code correlation preprocessing module described above processes homogeneous patent classification codes. Hereinafter, the heterogeneous multi-patent classification code correlation pre-processing module will be described. The heterogeneous multi-patent classification code correlation pre^¬ processing module has nearly the same operation principle as that of the homogeneous multi-patent classification code correlation pre-processing module. Almost all of patent documents issued in the United States include both IPC codes and USPC codes, and patent documents issued in Japan also include both IPC codes and FT codes. Let us assume that one patent document is granted IPCl, ..., IPCn and USPCl, ..., USPCm. Then, in view of the heterogeneous multi-patent classification code correlation processing, there are four types of correlation combinations between IPCl, ..., IPCn and USPCl, ..., USPCm. First, the correlation combination between IPCl and USPCl corresponds to a couple of a main IPC code and a main USPC code. Second, the correlation combination between IPCl and each of USPC2,...,USPCm corresponds to a couple of a main IPC code and a sub USPC code. Third, the correlation combination between each of IPC2,... IPCn and USPCl corresponds to a couple of a sub IPC code and a main USPC code. Fourth, the correlation combination between each of IPC2,... IPCn and each of USPC2, ...,USPCm corresponds to a couple of a sub IPC code and a sub USPC code.

*610For the heterogeneous multiple patent classification codes of each of the four types, the heterogeneous multi-patent classification code correlation pre-processing module can perform the same operation as the homogeneous multi-patent classification code correlation pre-processing module. At this time, since the processed codes are heterogeneous patent classification codes, they have no minimum common patent classification code and can be limited to preset values. For example, a couple of a given IPC code and a USPC code may be limited to have higher patent classification codes up to the subclass of the IPC and the class of the USPC. That is, as the homogeneous multi-patent classification code correlation pre-processing module generates the above-described correlation tables for a given patent classification code combination (Ai, Bj) (wherein Ai and Bj are patent classification codes of the same system), the heterogeneous multi-patent classification code correlation pre- processing module may generate correlation tables as described above for a given patent classification code combination (Ai, Bj) (wherein Ai and Bj are patent classification codes of different systems). The generated correlation tables may be granted a weight or contribution degree weight. Further, it is obvious to one skilled in the art that it is possible to make each heterogeneous patent classification code combination (Ai, Bj) (wherein Ai and Bj are patent classification codes of different systems) correspond to a patent document number and is possible to perform counting and calculation reflecting bibliographical information obtained from the patent document number. In addition, it is possible to acquire a heterogeneous patent classification code combination (Ai, Bj) for a particular document set and perform the same analysis for the combination as the homogeneous multi-patent classification code correlation preprocessing module.

Patent classification code-based statistic pre-processing module Hierarchy of patent classification code

A patent classification code-based statistic pre-processing module of the present invention performs a pre-processing for finding a preset statistic value for patent classification codes in at least one patent classification code system from at least one nation-based patent DB.

FIG. 61 illustrates a structure of the patent classification code- based statistic pre-processing module. The patent classification code- based statistic pre-processing module includes a patent classification code-based statistic pre-processing engine 3210 for pre-processing statistic values according to the patent classification code, and a pre- processed patent classification code-based statistic DB 3230 corresponding to a view, table, or information system storing the pre-processed patent classification code-based statistic values.

Patent classification codes dealt with the patent classification code-based statistic pre-processing module include at least one of the IPC, USPC, FT, FI, and ECLA codes. It is obvious that nation-based patent DBs processed by the patent classification code-based statistic pre-processing module include patent DBs issued by Republic of Korea, the United States of America, Japan, and Europe(EPO), as well as patent DBs issued by the patent offices of the other nations. Each nation may commonly use the IPC, while having its own specific classification system (e.g. US patent office has USPC or UPC, Japanese patent office has FT or FI, and Europe patent office has ECLA). In finding a preset statistic value from a patent DB of a nation having two or more patent classification systems, it is preferable to independently perform the searching job for each of the two or more patent classification systems. For example, in the case of US, it is necessary to perform the pre-processing for the IPC and the USPC, respectively. Further, each patent classification code system has its own levels. Specifically, the IPC system has levels including section, subsection, class, subclass, group, and subgroup, and has sub-levels under the subgroup, which have a hierarchical structure expressed by the number of dots attached to the title information corresponding to a relevant patent classification code. An example of the hierarchical structure has been described in detail above in relation to the patent classification code pre-processing module 301-3-1 or 3500. Therefore, levels of IPC patent classification codes according to the present invention include IPC section, class, subclass, main group, ldot subgroup, 2 dot subgroup, ..., and n-dot subgroup. Referring to those levels, it is possible to identify their hierarchical structure in the classification code up to the main group. However, it is impossible to identify their hierarchical structure from the subgroup without the dot information included in the title. Use of the number of dots in identifying the hierarchical structure is employed from the subclass in the case of the USPC system and after preceding seven digits (theme code + two English characters) in the entire FT codes in the case of the FT system.

FIG. 100 is a flowchart showing an example of pre-processing of statistic values according to each patent classification code by the patent classification code-based statistic pre-processing module. The patent classification code-based statistic pre-processing module modifies and stores patent classification codes in a database by using a patent classification code tree structure, so that a particular patent classification code system of a first nation may automatically include classification codes of a given patent classification code (S5020), obtains a search formula or search query including a patent classification code (S5030), converts a given patent classification code included in the search formula or search query to a modified patent classification code (S5040), performs a search by using a search formula or search query including the modified patent classification code by a search engine or DBMS 201 (S5050), obtains patent document information as a result of the search and modifies the obtained patent document information to be measurable (S5060), and performs a function of counting rank information of the result of the search in the unit of preset time or synthesized time based on a preset statistic value, parameter, or calculation value for a subject, such as applicant, inventor, or agent (S5070).

^Further, the patent classification code-based statistic preprocessing module can generate a statistic value, parameter or calculation value, which automatically includes lower patent classification codes of a given patent classification code, and an example of such a method is illustrated in FIG. 101.

The patent classification code-based statistic pre-processing module obtains a search formula including a patent classification code (S5120), obtains relevant patent document information including patent information lower patent classification codes of the patent classification code included in the search formula (S5130), and processes backward citation information and forward citation information included in the patent document, so as to generate a preset statistic value, parameter, or calculation value (S5140).

Further, the patent classification code-based statistic pre- processing module can generate a patent classification code-based statistic value, parameter, or calculation value, which reflects backward citation information or forward citation information including lower patent classification codes, for a subset of a particular document set, and an example of such a method is illustrated in FIG. 102. The patent classification code-based statistic pre-processing module obtains at least one document set (S5220), extracts information on the applicant, inventor, agent, at least one patent classification code of every document included in the document set (S5230), generates a document subset from the document set based on at least one item of the extracted information (S5240), obtains information on prior-filed documents cited by all documents of the document subset (S5250) or obtains information on later-filed documents citing all the documents of the document subset (S5260), and generates a preset backward citation and forward citation- relating statistic value, parameter or calculation value for the obtained prior-filed documents and/or later-filed documents (S5270).

Hereinafter, the present invention≤] patent classification code- based statistic pre-processing module will be described in more detail in view of the patent classification code-based statistic pre-processing engine.

Options

When the patent classification code-based statistic pre-processing module processes the preset statistic value, the following options are taken into consideration and can be selected in a crossed manner.

First, when each patent classification code is given, whether information on lower classification codes of the given patent classification code will be automatically included in the information processed by the module is in question. This issue is more important especially in a level in which it is necessary to identify the hierarchical structure from the number of dots attached to the title information. For example, although it is possible to easily find, by using an extension, a patent document including a patent classification code under a given IPC level up to the levels of H, H04, H04B, and H04B 7/00, it is impossible to find, without using a lower classification- including search method of the present invention, a patent document including a patent classification code belonging to the lower classification in the subgroup level. For example, for a given IPC code of H04B 7/15, a statistic value without lower classification codes can be generated only by obtaining patent documents including H04B 7/15 from among patent documents of each nation. However, in order to generate a statistic value including lower classification codes, it is required to obtain only the patent documents including H04B 7/15 but also patent documents including H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216 under H04B 7/15 in the IPC hierarchical structure. The same can be said about the USPC, FT, etc., and the same description thereof is omitted here.

In the case of processing the statistic value automatically including lower classification codes, a range search method (including both use of a search engine and use of a DBMS 201) according to the present invention may be utilized. Meanwhile, in order to achieve the object of the present invention, further to the range search, a given patent classification code can be processed by another method. That is, it is possible to first store all lower patent classification codes of the given patent classification code, and then process a statistic value of the given patent classification code together with statistic values of all the lower patent classification codes of the given patent classification code, instead of processing only the statistic value of the given patent classification code alone. For example, in the case of H04B 7/15 described above, it is possible to first obtain information that H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216 exist under H04B 7/15 from the patent classification code hierarchical structure information, store the existence of H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216 under H04B 7/15, and obtain a preset statistic value automatically including H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216 existing under H04B 7/15 in response to a command requiring the preset statistic value for H04B 7/15. The same can be applied to other patent classification code systems, such as USPC and FT.

For a patent classification code of a given level, it is preferable to generate a preset statistic value automatically including lower patent classification codes of the given patent classification code, rather than simply generating a preset statistic value for only documents including the given patent classification code.

The second issue is how to deal with a case in which there exist multiple patent classification codes. Countermeasures for this case may include^ 1) a method of employing only the main patent classification code (usually, the first appearing code is the main patent classification code) as an object of the statistic processing; 2) a processing method giving a uniform weight to the main patent classification code and sub-patent classification codes! 3) and a processing method giving different weights to the main patent classification code and sub-patent classification codes (for example, when there are (n+1) patent classification codes, the main patent classification code may be granted a weight of 50% while the sub- patent classification codes are granted a weight of 50%*l/n). Among the three methods 1) to 3), methods 1) and 3) are preferred, and method 3) is most preferred.

The third issue is a processing method when there are multiple applicants or multiple inventors. Countermeasures for this case include: 1) a method of giving a weight of 100% to each applicant or inventor and 2) a method of equally dividing the weight by the number of applicants or inventors (when there are n inventors, each inventor is granted a weight of 100%*1/n). It is preferred that each applicant is granted a weight of 100% in view of independence of implementation, while each inventor is granted a weight of 1/n. The same method as that for the applicant or inventor can be applied to the agent, and it is preferred that each agent is given a weight of 100%.

For a given patent classification code, the patent classification code-based statistic pre-processing module either may count in advance either the number of just lower patent classification codes of the given patent classification code in the patent classification code system, or the number of all the lower patent classification codes of the given patent classification code.

For a given particular patent classification code, the patent classification code-based statistic pre-processing module may calculate in advance the following information based on application and/or registration according to preset terms in a predetermined nation-based patent DB. The above-mentioned calculation can be achieved because it is possible to determine a patent document set satisfying the conditions, such as the patent classification code, appointed nation, preset time, and application or registration. By counting for the determined patent document set, it is possible to calculate various parameters as follows.

For example, based on the application, it is possible to calculate in advance the number of applications based on the year or quarter, application number increase rate, application number increase speed, and application number increase acceleration. The each-year's application number can be calculated by obtaining the number of all patent documents filed in a particular year. The increase rate can be calculated by (current term number past term number)/past term number*100%, the increase speed can be calculated by (current term number past term numberVtime interval, and the increase acceleration can be calculated by a change rate with respect to the time interval of the increase speed.

Further, the number of applicants of each year can be obtained (in the case of a joint application, the above-mentioned options can be applied). Once the number of applicants is calculated, the applicant number increase rate can be calculated, and an average applicant number and an average applicant number increase rate for each application can also be calculated. Further, an each-year's inventor number, inventor number increase rate, average inventor number, and average inventor number increase rate can be calculated by the same method as that of the each- year's applicant number. Also, since it is possible to calculate the number of claims in each application, it is also possible to obtain an each-year's application claim number, each-year's application claim number increase rate, average application claim number of each application, average application claim number increase rate of each application. Here, since the application claims can be divided into independent claims and dependent claims, it is possible to calculate the number of independent claims, the number of dependent claims, and their increase rates.

Further, by calculating the number of patent classification codes of each application, it is possible to obtain a each-year's patent classification code number and patent classification code number increase rate and obtain an average patent classification code number and average patent classification code number increase rate for each application. Further, in the case of a patent DB of a nation using multiple patent classification codes, such as US or Japan, those values can be calculated in advance for each patent classification code. When each application has family information of the application, it is possible to calculate an each-year's family number, an each-year's family number increase rate, and an average family nation number and an average family nation number increase rate of each application based on the family information.

Although the calculation described above is based on the application, the same parameters can be calculated based on the registration. Parameters calculable based on the registration include registration number, registration number increase rate, registration number increase speed, registration number increase acceleration, registered party number, registered party number increase rate, average registered party number, average registered party number increase rate, inventor number, inventor number increase rate, average inventor number, average inventor number increase rate, registered claim number, registered claim number increase rate, average registered claim number, average registered claim number increase rate, patent classification code number, patent classification code number increase rate, average patent classification code number, average patent classification code number increase rate, family number, family number increase rate, average family nation number, and average family nation number increase rate.

Those parameters enumerated above are only an example, and it is possible to calculate various other parameters from various types of numerical information included in the application or registration patent information and previously counted values of the present invention. The parameters described above essentially include counted values, change rates and change values of the counted values, statistic quantities (average, standard deviation, etc.) of the counted values, and function values (e.g. registration rate) of at least two counted values.

Discovery of quantity-based subject The patent classification code-based statistic pre-processing module may previously calculate the following information according to each of preset terms based on the application and/or registration, for patent DB of each nation in relation to a given particular patent classification code. Such a calculation is possible because it is possible to determine a patent document set satisfying conditions including the patent classification code, appointed nation, preset time, and application or registration. By counting for the determined patent document set, it is possible to calculate various parameters as follows. For example, through counting for the determined patent document set, it is possible to extract each-year's or whole-year's maximum applicants, inventors, agents and give rankings to the extracted applicants, inventors, or agents.

Further, it is possible to calculate a concentration degree, an activity index, etc, of each of the extracted applicants, inventors, or agents, and to obtain rankings of the applicants, inventors or agents based on the calculated concentration degree or to obtain rankings of the applicants, inventors or agents based on the calculated activity index. Here, the applications-number-based concentration degree can be calculated by (all applications of a particular subject having a particular patent classification code during a particular period/all applications of the particular subject during the particular period)*100%. Further, the activity index can be calculated by [(all applications of a particular subject having a particular patent classification code during a particular period/all applications of the particular subject during the particular period)/(all applications having the particular patent classification code during the particular period/all applications during the particular period)*100%]. Here, since it is possible to obtain, by a search engine or the DBMS 201, 1) all applications of a particular subject having a particular patent classification code during a particular period, 2) all applications of the particular subject during the particular period, 3) all applications having the particular patent classification code during the particular period, and 4) all applications during the particular period, it is possible to calculate the concentration degree or activity index. Meanwhile, the concentration degree or activity index is only an example of patent indexes. The search engine or DBMS 201 can calculate values of functions, input values of which are numerical values that can be obtained either indirectly through pre-calculation or directly, and can give rankings to the applicants, inventors, or agents based on the calculated function values. The function values according to the present invention as described above may include all patent analysis indexes defined as values obtainable by the search engine or DBMS 201in a patent DB from among all conventional definable patent analysis indexes that can be utilized as patent analysis indexes. Further, instead of calculating the rankings of applicants, inventors or agents based on the number of applications as described above, it is possible to calculate the rankings of applicants, inventors or agents for the patent index or function based on other parameters, such as applications-number increase rate and applications-number increase speed. Those parameters may include the number of application claims, application claim number increase rate, each-application's average application claim number, each-application's average application claim number increase rate, patent classification code number, patent classification code number increase rate, each application's average patent classification code number, each application's average patent classification code number increase rate, family number, family number increase rate, each application's average family nation number, and each application's average family nation number increase rate.

Although various patent indexes described above are based on the application, it is obvious to one skilled in the art that they can also be defined and calculated based on the registration, because the selection between the application and the registration is a problem of specification or determination of a patent document set. Therefore, it is possible to specify or determine a patent document set from only applications having a request for examination in the application stage. Further, in the case of employing a utility model system, an issue whether the application or registration includes a utility model application or a utility model registration is also a problem relating to the specification or determination of the patent document set, and there are various methods and standards for the specification or determination in the level of one ski lled in the art .

Citation

The patent classification code-based statistic pre-processing module can generate citation-related information as follows based on the application and/or registration for each preset term from a nation-based patent DB (e.g. a patent DB constructed by patent documents issued by the US patent office)including citation information of a given particular patent classification code. Such a calculation is possible because it is possible to determine a patent document set satisfying the conditions, such as the patent classification code, appointed nation, preset time, and application or registration. For example, if USPC=002/456 (Body cover in GUARD OR PROTECTOR in MISCELLANEOUS in apparel)is given, it is possible to configure a patent document set including USPCO02/456 as a patent classification code of its own patent document. Through counting in relation to the determined patent document set, it is possible to various parameters as follows.

Bibliographical information of a US patent document includes citation information, such as others' patents cited by the document itself. That is, when document B cites document A and document a, document A and document a are cited documents of document B, and document B is a citing document of document A and document a. In this case, an applicant (including assignee), an inventor, an agent, and a patent classification code relating to document B are a citing applicant, a citing inventor, a citing agent, and a citing patent classification code, respectively, and an applicant (including assignee), an inventor, an agent, and a patent classification code relating to document A and document a are a cited applicant, a cited inventor, a cited agent, and a cited patent classification code, respectively. When document B cites document A and document a, bibliographical information of document B contains numbers of document A and document a. When document B cites document A and document a, there are a mapping relation of A->B and a->Band a mapping relation of B->A and B->a in view of citation in the computer technology, and the mapping relation can establish the correlation of A<->B and a<->B. Therefore, documents cited by document B can be easily constructed as data included in the bibliographical information of document B, and it is possible through the mapping to easily specify a document set citing document Aand a document set citing document a. That is, both the document set citing document a and the document set citing document a are expected to include document B.

This process can be performed for all the determined or specified documents. That is, when all specified patent documents are processed one by one, it is possible to establish mapping relations between all patent documents through a computational method or other methods. When the documents are processed one by one, it is natural that the document, which is an object of each processing, is in the standpoint of document B. At this time, when both document A and document a are included in a universal set named US patent DB, information in relation to document A and document a may include a record of mapping information between documents A and a and document B. When one of document A and document a is a document managed b a patent DB of another nation also, mapping information between documents A and a and document B can be recorded in information relating to the document of the patent DB of that nation.

When a patent document set including citing information (a document set including document B) is specified, the patent classification code-based statistic pre-processing module can collect cited patent document numbers (numbers of documents A and documents a) included in patent documents of the patent document set, wherein the collected cited patent document number has a corresponding cited patent document, and the cited patent document naturally includes bibliographical information. Therefore, a cited patent document set cited by all the patent documents of the specified patent document set can also be specified at a particular calculation time point. Accordingly, the specified cited patent document set can be an object of analysis, counting or calculation. Here, repetitive counting is naturally permitted. That is, when one cited document is cited by multiple documents of the patent document set including the citing information, the cited document is required to have a weight or counting value corresponding to the number of its citations. That is, it is rational to consider that each cited document number has been multiplied by a weight corresponding to the number of its citations. For example, when document A has been cited five times and document a has been cited three times, it is preferred to give a weight of 5 to document A and a weight of 3 to document a in calculation of the following parameters, such as the number of all citing applications. Based on the patent document set including the citing information, a word j°citingj± can be attached to all calculation values for the cited patent document set. In the repetitive counting, it is possible to calculate a ranking of a cited document number, which has been cited the largest number of times.

The patent classification code-based statistic pre-processing module can calculate the following information for the cited patent document set (a document set including documents A and documents a) or a specified patent document set. Further, as quantitative standards, it is possible to obtain the number of citing applications, citing applications- number increase rate, and citing applications-number increase speed (these can be understood from the number of documents included in the cited patent document set), and it is possible to calculate the citing applications-number increase acceleration, citing applicant number, citing applicant number increase rate, citing average applicant number, citing average applicant number increase rate, citing inventor number, citing inventor number increase rate, citing average inventor number, citing average inventor number increase rate, citing application claim number, citing application claim number increase rate, citing average application claim number, citing average application claim number increase rate, citing patent classification code number, citing patent classification code number increase rate, citing average patent classification code number, and citing average patent classification code number increase rate. The method of counting the citing applications-number, citing application number increase rate, citing application number increase speed, citing application number increase acceleration, citing applicant number, citing applicant number increase rate, citing average applicant number, citing average applicant number increase rate, citing inventor number, citing inventor number increase rate, citing average inventor number, citing average inventor number increase rate, citing application claim number, citing application claim number increase rate, citing average application claim number, citing average application claim number increase rate, citing patent classification code number, citing patent classification code number increase rate, citing average patent classification code number, and citing average patent classification code number increase rate is the same as the method of counting the parameters as described above. That is, if the number of applications, applications-number increase rate, applications-number increase speed, application number increase acceleration, applicant number, applicant number increase rate, average applicant number, average applicant number increase rate, inventor number, inventor number increase rate, average inventor number, average inventor number increase rate, application claim number, application claim number increase rate, average application claim number, average application claim number increase rate, patent classification code number, patent classification code number increase rate, average patent classification code number, and average patent classification code number increase rate are counted from the cited patent document set, obtained values correspond to the number of the citing applications, citing applications-number increase rate, citing applications-number increase speed, citing applications-number increase acceleration, citing applicant number, citing applicant number increase rate, citing average applicant number, citing average applicant number increase rate, citing inventor number, citing inventor number increase rate, citing average inventor number, citing average inventor number increase rate, citing application claim number, citing application claim number increase rate, citing average application claim number, citing average application claim number increase rate, citing patent classification code number, citing patent classification code number increase rate, citing average patent classification code number, and citing average patent classification code number increase rate, respectively.

In the same way, the patent classification code-based statistic pre-processing module can calculate rankings of applicants, inventors, agents or patent classification codes according to each of the parameters from the cited patent document set. That is, the patent classification code-based statistic pre-processing module can calculate rankings of maximum cited applicants and rankings of maximum cited inventors based on the number of applications. In calculating the rankings of the patent classification codes, the patent classification code-based statistic pre^¬ processing module can calculate rankings of maximum cited patent classification codes for each stage in each patent classification code system by using a lower patent classification code system of the present invention. That is, it is natural to consider that each patent document included in the cited patent document set includes at least one patent classification code and each of all higher patent classification codes of the included patent classification code (in the patent classification code system) also has been cited. Therefore, it is preferred that this point of view is reflected in calculation of the each-stage maximum cited patent classification code rankings. US patent documents included in the cited patent document set may include patent classification codes of both USPC and IPC.

The patent classification code-based statistic pre-processing module can acquire applicant information from a patent document set including the citing information (a document set including document B), can specify the cited patent document set from an applicant-based patent document set including the citing information (a patent document set relating to one of all the applicants is expected to include document B), and can calculate various parameters, such as the number of citing applications, from the specified cited patent document set. At this time, for each parameter also, the rankings of the applicants can be calculated. For example, the patent classification code-based statistic pre-processing module can extract applicants from a patent document set of each year specified by USPC=002/456 (Body cover in GUARD OR PROTECTOR in MISCELLANEOUS in apparel), and generate a document set for each of the extracted applicants (e.g. for US registration No. 06401262, assigned to Benetton Group S.p.A., US Cl. 2/456; 2/411, the patent classification code-based statistic pre-processing module can generate a document set corresponding to USPC=002/456for the Benetton Group S.p.A.). The patent classification code-based statistic pre-processing module can generate the cited document set from the each-applicant-based document set in the same way and can calculate various parameters for the cited document set.

In view of mapping, the patent classification code-based statistic pre-processing module can generate a citing document set, i.e. a document set citing each patent document included in the specified patent document set. For example, when document B is cited by document C and document c, it is noted from the mapping relation that document B is cited by document C and document c. In order to find document C and document c, the number of document B is input in the citing document number field of a search engine. Then, document C, etc. having citing information including document B is output as a result of the search. Of course, by using a query, such as select, while appointing the number of document B as a citing number field in the DBMS 201, it is possible to obtain document C, etc. as a result of the search. If the mapping relation of B->C and B->c has been arranged as data, it is possible to easily specify a document set including document C citing document B.

Therefore, the patent classification code-based statistic preprocessing module can calculate the following cited-by information (in view of the cited document B, information on the documents included in the citing document C is cited-by information) from a citing document set citing each patent document of the specified patent document set.

Once the patent document set (document set including document) is specified, the patent classification code-based statistic pre-processing module can collect numbers of citing patent document citing each patent document of the specified patent document set (number of document C and number of document c) by making an inquiry to the search engine or DBMS 201. It goes without saying that the number of each of the collected citing patent document corresponds to bibliographical information. Therefore, a citing patent document set citing all patent documents of the specified patent document set is also specified at a particular calculation time point. Therefore, the specified citing patent document set can be an object of analysis, counting or calculation.

At this time, it is preferable to allow repeated counting. When the specified patent document set includes document B and document b and when both document B and document B are cited by document C, it is natural that document C should be given a weight of 2. That is, document C belongs to both a result of search by document B and a result of search by document b. Based on the specified patent document set, all calculation values for the citing patent document set may be given a name "cited" or "cited- by." In the case of repeated counting, it is possible to calculate rankings of citing document numbers most-frequent Iy citing the patent documents belonging to the specified patent document set. The patent classification code-based statistic pre-processing module can obtain the above-mentioned information for the citing patent document set (document set including document C and document c).

In the same way, from the citing patent document set, the patent classification code-based statistic pre-processing module can calculate rankings of applicants, inventors, agents or patent classification codes according to each parameter. That is, the patent classification code- based statistic pre-processing module can calculate rankings of maximum citing applicants and rankings of maximum citing inventors based on the number of applications. In calculating the patent classification code- based rankings, it is possible to calculate rankings of maximum cited patent classification codes for each stage of each patent classification code system by using the lower patent classification code system of the present invention. That is, it is natural to consider that each patent document included in the citing patent document set includes at least one patent classification code and each of all higher patent classification codes of the included patent classification code (in the patent classification code system) also relates to the citing. Therefore, it is preferred that this point of view is reflected in calculation of the each- stage maximum citing patent classification code rankings. US patent documents included in the citing patent document set may include patent classification codes of both USPC and IPC.

The patent classification code-based statistic pre-processing module can analyze the specified patent document set (which includes document B and document b) itself. At this time also, it is preferred to grant a weight corresponding to the number of repetition by the allowed repetitive counting. For example, when document B has been cited five times and document b has been cited three times, it is preferred to give a weight of 5 to document B and a weight of 3 to document b in calculation of the following parameters, such as the number of all cited applications. By reflecting the weights described above, the patent classification code- based statistic pre-processing module can calculate rankings of most- frequent Iy cited patent documents in the specified patent document set. Further, first, in view of the quantity, the patent classification code- based statistic pre-processing module can count or calculate the number of cited applications (since document B is cited by document C, document B can use the name j°citedj±) from the specified patent document set (which includes document B and document b).

The patent classification code-based statistic pre-processing module can acquire applicant information from the specified patent document set (document set including document B), can specify the citing patent document set from the applicant-based patent document set (document B is expected to be included in a patent document set of one applicant from among the most-frequent applicants), and can calculate various parameters, such as the number of cited applications, from the specified citing patent document set. At this time, for each parameter, the rankings of the applicants can be calculated. For example, the patent classification code-based statistic pre-processing module can extract applicants from a patent document set of each year specified by USPC=002/456 (Body cover in GUARD OR PROTECTOR in MISCELLANEOUS in apparel), and generate a document set for each of the extracted applicants (e.g. for a document of US registration No. 06401262, assigned to Benetton Group S.p.A., US Cl. 2/456; 2/411, the patent classification code-based statistic pre-processing module can generate a document set corresponding to USPC=002/456for the Benetton Group S.p.A.). The patent classification code-based statistic pre-processing module can generate the citing document set from the each-applicant-based document set in the same way and can calculate various parameters for the citing document set.

The method of calculating various parameters by the patent classification code-based statistic pre-processing module of the present invention is characterized in that calculation values are generated for each specified or determined patent document set. Therefore, in a standpoint of the patent classification code-based statistic pre^¬ processing module, existence of a specified patent document set itself is important and attributes and size of the patent document set are not the problem (however, in order to perform calculation in relation to citing, the specified patent document set is required to include citing information).

Further, the patent classification code-based statistic preprocessing module of the present invention can calculate at least one parameter among the parameters for a patent document set appointed by a user using the patent classification code-based statistic pre-processing module.

Patent information processing basic module 40 Hereinafter, a patent information processing basic module 40 of the present invention will be described. The patent information processing basic module 40 includes 1) a search engine module, 2) a multidimensional analysis operation result table generation module 402, 3) an analysis module, 4) a monitor module 403, 5) a patent document set obtaining module, 6) a directory creation module 405, 7) a reporting module 406, 8) simplified analysis module 407.

Heterogeneous multi-patent classification code correlation preprocessing module The homogeneous multi-patent classification code correlation pre^¬ processing module described above processes homogeneous patent classification codes. Hereinafter, the heterogeneous multi-patent classification code correlation pre-processing module will be described. The heterogeneous multi-patent classification code correlation pre- processing module has nearly the same operation principle as that of the homogeneous multi-patent classification code correlation pre-processing module. Almost all of patent documents issued in the United States include both IPC codes and USPC codes, and patent documents issued in Japan also include both IPC codes and FT codes. Let us assume that one patent document is granted IPCl, ..., IPCn and USPCl, ..., USPCm.

Then, in view of the heterogeneous multi-patent classification code correlation processing, there are four types of correlation combinations between IPCl, ..., IPCnand USPCl, ..., USPCm. First, the correlation combination between IPCl and USPCl corresponds to a couple of a main IPC code and a main USPC code. Second, the correlation combination between IPCl and each of USPC2,...,USPCm corresponds to a couple of a main IPC code and a sub USPC code. Third, the correlation combination between each of IPC2,... IPCn and USPCl corresponds to a couple of a sub IPC code and a main USPC code. Fourth, the correlation combination between each of IPC2,... IPCn and each of USPC2,...,USPCm corresponds to a couple of a sub IPC code and a sub USPC code.

For the heterogeneous multiple patent classification codes of each of the four types, the heterogeneous multi-patent classification code correlation pre-processing module can perform the same operation as the homogeneous multi-patent classification code correlation pre-processing module. At this time, since the processed codes are heterogeneous patent classification codes, they have no minimum common patent classification code and can be limited to preset values. For example, a couple of a given IPC code and a USPC code may be limited to have higher patent classification codes up to the subclass of the IPC and the class of the USPC. That is, as the homogeneous multi-patent classification code correlation pre-processing module generates the above-described correlation tables for a given patent classification code combination (Ai, Bj) (wherein Ai and Bj are patent classification codes of the same system), the heterogeneous multi-patent classification code correlation preprocessing module may generate correlation tables as described above for a given patent classification code combination (Ai, Bj) (wherein Ai and Bj are patent classification codes of different systems). The generated correlation tables may be granted a weight or contribution degree weight. Further, it is obvious to one skilled in the art that it is possible to make each heterogeneous patent classification code combination (Ai, Bj) (wherein Ai and Bj are patent classification codes of different systems) correspond to a patent document number and is possible to perform counting and calculation reflecting bibliographical information obtained from the patent document number. In addition, it is possible to acquire a heterogeneous patent classification code combination (Ai, Bj) for a particular document set and perform the same analysis for the combination as the homogeneous multi-patent classification code correlation pre^¬ processing module. Patent classification code search module(401)

Hereinafter, the patent classification code search module 401 will be described. Search of a patent classification code refers to a search process in which, when a technical keyword is input as a search phrase, a technical classification code including the technical keyword as title information is output as a search result.

Hereinafter, a patent classification code search according to the present invention is described based on the following example.

Section : H electricity Class : HOl basic electrical element Sub-class : HOF magnet main group: HOlF 1/OOmagnet or magnetic body characterized by its magnetic material

1-dot sub-group : HOlF 1/01 *those made from inorganic materials 2-dot sub-group : HOlF 1/03 **those characterized by coercive force

3-dot sub-group : HOF 1/032 *** those made from hard magnetic material

4-dot sub-group : HOF 1/04 **** metal or alloy 5-dot sub-group : HOF 1/047 ***** alloy characterized by composition

6-dot sub-group : HOF 1/053 ******those including rare-earth metal

*696The first characteristic of the patent classification code search according to the present invention is to provide a lower patent classification code found through the search, together with higher patent classification codes of the lower patent classification code found through the search. By way of an example, when "hard magnetic material" is input as a search phrase, it is preferable to present "HOF 1/032: those made from hard magnetic material" as a result of the search, simultaneously while presenting higher patent classification codes of HOF 1/032 up to a preset level in the patent classification code system tree as a result of the search, wherein the higher patent classification codes of HOlF l/032are obtained by referring to the patent classification code master DB 203 for HOlF 1/032. That is, it is preferred that, when HOF 1/032 is input as a search phrase, the following result is output as a search result. It is preferred that the preset level is subclass in the case of IPC, class in the case of USPC, theme in the case of FT, and subclass in the case of ECLA or FI. Of course, it is okay to present a search result even above or below the preset level .

Sub-class : HOlF magnet main group: HOlF 1/00 magnet or magnetic body characterized by its magnetic material

1-dot sub-group :H01F 1/01 *those made from inorganic materials 2-dot sub-group: HOF 1/03 **those characterized by coercive force

3-dot sub-group :H0F 1/032 *** those made from hard magnetic material

In order to output the results as described above, the following stages are necessary. First, a keyword input as a search phrase is searched for in the patent classification code indexes including patent classification codes and title information attached to the patent classification codes, so as to find at least one patent classification code as a result of the search, which includes the input search phrase in the title information of the patent classification code (when "hard magnetic material" is input, HOF 1/032 is found). Second, the found patent classification code is searched for in the patent classification code master DB 203, so as to find higher patent classification codes of the patent classification code up to a preset stage ( HOF 1/03, HOF 1/01, HOF 1/00, and HOF are found). Third, the found higher patent classification codes are searched for in the patent classification code master DB 203, so as to obtain title information of the found higher patent classification codes. Fourth, patent classification codes and their title information obtained in the first to third stages described above are output or displayed together with information indicating the hierarchical structure including dots.

Meanwhile, when "hard magnetic material" AND "rare-earth" is input as a search phrase in the patent classification code search, it is impossible to yield a result of the search because there is no title information including both the "hard magnetic material" and the "rare- earth." However, in consideration of the patent classification code system, it is required that "1/053 ******those including rare-earth metal" should be output as a result of the search. That is, the simple line-by^¬ line indexing as described above cannot provide a satisfactory search result reflecting the systematic property of patent classification codes. In order to solve the above-mentioned problems, it is necessary to modify the patent classification codes and their title information in the following manner. Essence of the modification is to make title information of each patent classification code be incorporated with title information of its higher patent classification codes. Table 23 below shows an example of the modification.

[Table 23]

Although title information of each patent classification code includes title information of higher patent classification codes up to the highest patent classification code in Table 23, it is possible to allow title information of each patent classification code to be incorporated with title information of higher patent classification codes up to only a preset level. It is preferred that the preset level allowing the incorporation is subclass in the case of IPC, class in the case of USPC, theme subclass in the case of FT, and subclass in the case of ECLA or FI. In a state in which the patent classification codes have been indexed line by line with the incorporated title information, input of "hard magnetic material" AND "rare-earth" as a search phrase can yield "HOF 1/053" as a result of the search. Meanwhile, as described above, when one patent classification code has been yielded as a result of the search (for example, when HOF l/053has been yielded as a result of the search), it is possible to provide all or preset levels of higher patent classification codes of the found patent classification code together with their title information as a result of the search. That is, when "hard magnetic material" AND "rare-earth" has been input as a search phrase, the search result is required to include search data as follows.

6-dot sub-group :1/053 ******those including rare-earth metal

If "metal AND coercive force" is input as a search phrase, it is preferred that the following search result is finally output.

Sub-class : HOlFmagnet main group: HOF 1/OOmagnet or magnetic body characterized by its magnetic material 1-dot sub-group :H01F 1/01 *those made from inorganic materials

2-dot sub-group: HOlF 1/03 **those characterized by coercive force

Here, if a means (e.g. a (+) sign/mark), which enables extensional view of lower patent classification codes of HOlF 1/03, exists around HOlF 1/03 and a user has selected this means, it is preferable to hierarchically display the lower patent classification codes.

FIG. 16 illustrates an example of an operation of the patent classification code search module 401 according to the present invention. The search includes: a step for selection of a patent classification code a step for selection of a search phrase in order to find the patent classification code a step for input of at least one search phrase! and a step of performing the search using the input search formula, thereby generating a search result.

Meanwhile, by using the hierarchical structure (child-patent) information between the patent classification codes, the patent classification code search module 401 can achieve the object of the present invention without using the incorporated title information. When there are at least two patent classification codes obtained as a result of input of a search phrase/query phrase, it is possible to achieve the same object by examining if they have a child-parent relation in the patent classification code system.

Patent document set obtaining module 404

Hereinafter, a patent document set obtaining module 404 according to the present invention will be described. The patent document set obtaining module 404 includes an automatically-selected document set obtaining module 404-1 and a user-created document set obtaining module

404-2, and the user-created document set obtaining module 404-2 includes a search-formula-using document obtaining module 404-2-1 and a selection-on- directory document obtaining module 404-2-2 that enables obtainment of a document set through selection on a directory capable of specifying a document set, such as an IPC directory. Meanwhile, the patent document set obtaining module 404 can obtain or select a patent document set generated by either itself or others, and the patent document set obtaining module 404 can generate a new patent document set through a set operation (including a union, a difference set, a NOT operation) between at least two patent document sets. The search-formula-using document obtaining module 404-2-1 includes a search engine-using document set obtaining module 404-2-1-1 making an inquiry to the search engine, and a DBMS-using document set obtaining module 404-2-1-2 making an inquiry to the DBMS 201, which are divided according to an object to which the search formula is sent. The automatically-selected document set obtaining module 404-1 automatically obtain various determined patent document sets as described above in relation to the determination of the patent document set according to the present invention, except for the patent document sets generated by the user.

A multi-dimensional analysis operation result table generating module 402 Hereinafter, the multi-dimensional analysis operation result table generating module 402 of the present invention will be described. When data on patent information is configured in a DB, an SQL query is conventionally used to retrieve a required result. Herein, although a DB structure is we11-designed, multiple tables have to be joined to deduce one result. Also, when a table size is large, it takes a very long time to process a selection query.

For example, in obtaining each-year's occupancy rate of each numerously-filed IPC 1 dot subgroup of Korea patents of Samsung Electronics Inc. in the case of a select query, various types of tables which depend on the design of a DB schema but generally include several or more tables (for example, an applicant table, an IPC table, a document table, etc. of each country) have to be joined. First, in an IPC 1 dot subgroup level of Samsung Electronics Inc., a numerously-filed IPC has to be extracted, and herein, for automatic inclusion of a lower-level IPC, a lower-level IPC of each IPC has to be extracted. Then, for the extracted 1 dot subgroup IPC, based on the data of all Korea patents of Samsung Electronics Inc., the number of each-year's applications is searched out and numerously-filed 1 dot subgroup IPCs are searched out. Next, on each of the searched-out numerously-filed 1 dot subgroup IPCs, the ratio of the number of applications of Samsung Electronics Inc. to the number of all applications is obtained, thereby obtaining an occupancy rate of Samsung Electronics Inc. in each numerously-filed 1 dot subgroup IPC. If this is made into one SQL query, the SQL query is too long, and requires a long process time. Especially, in order to obtain the required result by automatically including a document corresponding to a lower-level patent classification code of the present invention, too many operations are required, which significantly reduces the response speed. Meanwhile, there is a possibility that variously modified requirements repeatedly occur, such as the same process for LG Electronics Inc., or the calculation of the occupancy rate of LG Electronics Inc. in an IPC subgroup level.

In this case, when there exists table data including a result from a multi-dimensional analysis operation in accordance with an analysis object, the operation being performed through a previous process of patent information in accordance with the variously modified requirements, it is possible to significantly improve the response speed by making a query in the table. The multi-dimensional operation includes a roll-up operation, a cube operation, or the like, and the result of such an operation for patent data is stored in a DB table. Herein, the table may be called a cube, a materialized view, a multi-dimensional analysis operation result table, a view, or the like, but the various names indicate the common subject. On the above described table, an SQL query is used to extract required data. In extracting the required data from the multi-dimensional result table of a cube operation, a MDX (multidimensional expression) query is conventionally used, but the MDX query is generally called SQL for convenience sake in the present specification, In general, an SQL query is made for access to a multi-dimensional analysis roll-up operation result table, and an MDX query is made for access to a multi-dimensional analysis cube operation result table. However, since both queries have a common essence of accessing a multi-dimensional analysis operation result table, and are almost identical from the standpoint of a query form, an "SQL query" is representatively used for convenience sake in the description of the present specification. In other words, an SQL query on a multi-dimensional analysis operation result table can be understood in two ways, such as 1) when the table made a roll-up operation, an SQL query is made, and 2) when the table made a cube operation, an MDX query is made. In other words, multi-dimensional analysis indicates that the analysis is performed by one or more dimensions (of course, one-dimensional analysis is included). In the multi-dimensional analysis operation result table of the present invention, an operation result of the steps from 1 to n-1 (or n) is stored when it is assumed that a user's required analysis result is logically generated by n information-processing steps. Herein, an operation result only up to an intermediate step is stored in the multi^¬ dimensional analysis operation result table, and a final analysis result may be generated by a DBMS or SQL query formula, because an SQL query or DBMS has an operation function.

The multi-dimensional analysis operation result table generating module 402 indicates an engine for performing a roll-up or cube operation of patent information, or another required operation or a data process, and generating the result as a table in order to rapidly output required information. Especially, an analysis module of the present invention, which uses the table, can rapidly generate required data by making a simplified SQJ- query in a table generated by the multi-dimensional analysis operation result table generating module 402 when a roll-up, a drill-down, or a drill-through occurs. For a patent document master DB 202, a patent classification code master DB 203, or a subject master DB 204 as an object, these multi^¬ dimensional analysis operation result table generating module 402 performs a roll-up operation and/or a cube operation in each of at least one or more analysis subject categories, and stores the results in a table. As described above, the patent document master DB 202 includes a bibliographical information master DB, etc. and the patent classification code master DB 203 includes a patent classification code DB according to each type of a patent classification code system, etc., and the subject master DB 204 includes an applicant name DB, an inventor name DB, etc. Herein, it is preferred that the applicant name DB includes a representative applicant name made for each nation.

In order to generate an analysis result in accordance with at least one analysis subject by data included in the patent document master DB 202, the patent classification code master DB 203, and the subject master DB 204, the multi-dimensional analysis operation result table generating module 402 performs a multi-dimensional analysis operation and generates the result in a table. The table may include multiple tables, and the multiple tables are configured as a DW (data warehouse). The DB schema of a patent information table used by the multi-dimensional analysis operation result table generating module 402 is preferably configured as a star-schema.

An E-R diagram (Entity Relationship Diagram) of tables re^¬ configured as a star-schema has a FACT table at the center, and a dimension table as a reference information table referring to the FACT table. The FACT table preferably includes bibliographical information data on a patent document in each field (the data of the dimension table is excluded, but in the dimension table, at least one ID of each dimension, such as an IPC_ID indicating the specific IFCXs) included in a patent document, an application ID indicating a specific apρlicant(s), a date ID indicating a date, is included). The dimension table includes tables of other objects that may be included in a dimension, such as a patent classification code table (IPC is necessary, and in the case of a FACT table of each nation, a patent classification code of the nation), a date table (for example, year), a subject table (for example, applicant/agent/inventor), a location table (for example, nation/location/company). Meanwhile, of course, in the dimension table, IDs corresponding to various IDs included in the FACT table are included. For example, an IPC dimension table includes the IPC_ID which corresponds to a specific one IPC. The tables re-configured as a star-schema may be basically configured according to each nation or unified nations. In unified nations or in other cases, some data related to multiple nations, such as family information data or INPADOC/DocDB data, is preferably managed by an additional table.

The multi-dimensional analysis operation result table generating module 402 generates a multi-dimensional analysis operation result table for various analysis objects of the present invention by using the data of a patent information table re-configured in a star-schema. Meanwhile, the multi-dimensional analysis operation result table generating module 402 may generate another multi-dimensional analysis operation result table by using the generated multi-dimensional analysis operation result table, so as to analyze another analysis subject or a complicated analysis subject. In other words, the multi-dimensional analysis operation result table generating module 402 uses at least one or more of 1) a patent information table re-configured in a star-schema, 2) a patent document master DB 202, a patent classification code master DB 203, and a subject master DB 204, and 3) a multi-dimensional analysis operation result table. Herein, for convenience sake, the patent information table re-configured in a star- schema configuration generally represents 1) to 3), but 2) and 3) are not excluded.

The relation between a multi-dimensional analysis operation result table generating module 402 and an analysis module

The multi-dimensional analysis operation result table generating module 402 includes an analysis DW generation module 402-1 for generating the analysis DW, various kinds of modules generating an analysis-subject- based multi-dimensional analysis operation result, and other modules for performing a function required for generating a multi-dimensional analysis operation result of the present specification. The various kinds of modules for generating an analysis-subject-based multi-dimensional analysis operation result include a total amount analysis multidimensional analysis operation result table generating module 402-2, a citation-analysis multi-dimensional analysis operation result table generating module 402-3, a competitive-analysis multi-dimensional analysis operation result table generating module 402-4, an inventor-analysis multi-dimensional analysis operation result table generating module 402-5, a patent technology classification-analysis multi-dimensional analysis operation result table generating module 402-6, a convergence-analysis multi-dimensional analysis operation result table generating module 402-7, a representing phrase-analysis multi-dimensional analysis operation result table generating module 402-8, etc. The modules generate an analysis DW 205-1, a total amount analysis multi-dimensional analysis operation result table 205-2, a citation-analysis multi-dimensional analysis operation result table 205-3, a competitive-analysis multi-dimensional analysis operation result table 205-4, an inventor-analysis multi-dimensional analysis operation result table 205-5, a patent technology classification- analysis multi-dimensional analysis operation result table 205-6, a convergence-analysis multi-dimensional analysis operation result table 205-7, and a representing phrase-analysis mult i-dimensional analysis operation result table 205-8, respectively. Hereinafter, the relation between the mult i-dimensional analysis operation result table generating module 402 and the analysis module of the present invention will be described. The analysis module includes at least one or more analysis formula (an SQL query) corresponding to an analysis subject, obtains a required analysis result of each analysis subject by the SQL query in the multi-dimensional analysis operation result table, processes the result provided from the system 1 of the present invention, and provides the processed result to a user of the present system.

Meanwhile, the multi-dimensional analysis operation result table generating module 402 may generate at least one or more tables according to each analysis object within a wider range than a predetermined scale/stage in consideration of the performance of the analysis module, such as decrease in extraction/calculation/obtaining time of the required data, and reduction of computer resource consumption. In other words, in the multi-dimensional analysis operation result table generating module 402, although final analysis result display data to be provided to a user's screen can be generated, it is more preferable that the final analysis result display data is generated by using various kinds of commands provided from an SQL query itself after data only up to certain intermediate stages in each analysis subject is generated. Herein, the latter case is preferable because lots of final analysis result displays result in huge inefficiency (waste of computer resource). However, if the optimization of computer resource is thrown up (the response speed is required to be maximized), the former case may be allowable. ROLL UP AND DRILL DOWN

Before processing patent information data into the physical view data, we will explain the concept of roll up and drill down. The roll up and drill down are basic functions in data computation, and they will be explained by an example in order for the persons in the art to understand the concept better. In Table 24, 2006' s application data are provided. [Table 24]

The shown data can be added into a quarter year unit and a year unit below. The following Table 25 is for the concept ional explanation, and it should be understood that the data structure of such roll up can be varied depening on actual development examples. [Table 25]

As described above, the roll up computation means adding from a small unit to a large unit in one direction with respect to one dimension at one time. That is, one example of the roll up computation is an addition from a month unit value to a quarter year unit value and then from the quarter year unit value to a year unit value, and thus the roll up means approaching, acquiring and extracting the result from the adding up to a higher unit in one dimension.

Meanwhile, 58 data in the year of 2006 are comprised of 6 cases in a first quarter, 15 cases in a second quarter, 24 cases in a third quarter, and 13 cases in a fourth quarter, and these 58 data in 2006 are again divided into 6 cases in a first quarter, 15 cases in a second quarter, 24 cases in a third quarter, and 13 cases in a fourth quarter (because they are added from the devided separate cases), and also cases in each quarter year can be divided into a month unit. As such, an operation of approaching, acquiring and extracting by subdividing from a bigger unit to a smaller unit in one dimension is called drill down. Therefore, the roll up and drill down is like both faces of a penny, and if there is a multi- dimension computed data by performing a multi-dimension computation such as a roll up computation, etc. on data from the smallest unit to the biggest unit, we can see the data by a drill down from the biggest higher rank unit to the lower rank unit.

Acquiring an objective data by approaching a multi-dimension analysis operation result table which is cube computed is typically called as OLAP (online analytic process), and a conventional OLAP engine supports functions such as a drill down, a drill through, etc. Roll up and Drill down

As seen from the following table, it can be appreciated that obtaining information on directly lower patent classification codes over a given patent classification code on a patent classification code system such as IPC and the like is important. In particular, the importance of an information processing method automatically including lower patent classification codes according to the present invention can be emphasized, wherein the method highlightes a necessity for acquiring the directly lower patent classification codes over a patent classification code (for example, a patent classification code below 1 dot subgroup for IPC) higher and lower of which can be distinguished by the recitation of a patent classification code.

The roll up and drill down will be explained as an example of table 26. The following table 33 represents the number of a patent application of a specific corporation "A" based on H04B 7/02. Numbers inside parentheses, ( ), refer to the number of documents classified as a specific IPC from among applied patent documents of corporation "A". Also, numbers inside braces, { }, are multi dimension computed numbers. Title information is added for reference, and it is independent of the roll up. However, because it is difficult for users to understand what a classification code means only by a patent classification code when drilling down, providing the title information is proper, and furthermore, the title information is included to facilitate understanding the multi- stage structure of the patent classification code (the reason why lower patent classification codes are included). 4 dot subgroup is not below H04B 7/02, but it is added for reference to show the multi-stage structure of the patent classification code. Except the below table, tables without the 4-dot subgroup and its title information will be provided and explained. [Table 26]

Let's have a look at H04B 7/02 to understand the roll up. A multidimensional computed number of H04B 7/02 equals 207 (=88+114+3+2), wherein 88 indicates the number of documents including H04B 7/02 from among patent documents, 114 is caused by H04B 7/02 and its lower patent classification, and 3 and 2 are caused by H04B 7/06 and H04B 7/08, respectively. Also, a multi-dimensional computed number of H04B 7/04 equals 114 (=57+36+21), wherein 57 indicates the number of documents including H04B 7/04 from among patent documents, and 36 and 21 are caused by H04B 7/06 and H04B 7/08, respectively. Also, multi-dimensional computed documents of H04B 7/00 equal 2650, which indicates the number of documents including H04B 7/00 from among patent documents and a multi-dimensional computed number in another patent classification code below this patent classification code. Herein, "..." included in Table 26 means that lower patent classification codes comprise other things except the recited in Table 33. It can be known that obtaining only information recited as H04B 7/02 is improper when patent document information for H04B 7/02 is taken (in search or counting, or any other statistic and analysis calculations). The reason is that it is more proper that the patent information for H04B 7/02 includes patent information for all patent classification code under H04B 7/02 in a hierarchical structure of a patent classification code system. (For example, if all documents in technical fields corresponding to H04B 7/02 are requested, it is proper that all documents relavant to all patent classification code under H04B 7/02 as well as documents recited as H04B 7/02 are output). Therefore, information processing (search or counting, or any other statistic and analysis calculations) including lower patent classification codes according to the present invention is needed. As can be seen from Table 33, if IPC information in a patent document has H04B 7/08, this document is relevant to H04B 7/04 and also is relevant to H04B 7/02. Therefore, from the point of view of counting such as the number of applications, the number of registrations and etc., a counting value for H04B 7/04 includes a value recited as H04B 7/04, a value recited as H04B 7/06 and a value recited as H04B 7/08.

It is well known that the above roll up and drill down is applicable to all levels of all IPC codes and all other patent classification codes. One example of drill down of each IPC will have a form below.

The following tables show one example of drilling down for each year. When drilling down information provided in the above Table 33 with respect to years, data such as the following Table 27 can be shown. For convenient recitation, a multi-dimension computed number is left alone but 4-dot sub-group having no data is not presented. [Table 27]

If H04B 7/00 is drilled down with respect to lower patent classification, data provided in the following Table 28 can be shown. [Table 28]

Herein, if drill down regarding H04B 7/015 is performed, H04B 7/015 does not have lower patent classification code information, so an additional change may not be found. Herein, if drill down regarding H04B 7/02 is performed, data provided in the following Table 29 can be shown. [Table 29]

Subsequently, if drill down regarding H04B 7/04 is performed, data provided in the following Table 30 can be shown. [Table 30]

The Table 27 through Table 37 show rolled up application date/application year for a specific patent classification code of corporation "A". That is, each multi-mension computed result can be divided into from the past to the year of 2000, 2001, 2002, 2003, 2004, 2005 and from the year in which a year and six months from the present date belongs to until the present date ("recently"is short for this period.) during drill down. To do this, a numerical value of documents corresponding to each classification code for a patent document set of corporation "A" needs to be rolled up in a year unit beforehand. The following Table 31 shows its one example.

For convenient description, yearly-based roll-ups and drill downs regarding H04B 7/02 will be explained. [Table 31]

For example, when a user performs an action such as clicking on (+) mark (for convenient description, this is a sign representing that drill down is possible) next to 2001, this system 1 can accept that user's action, extract data for lower dimensions (for example, a quarter) of the year of 2001 and provide it for the user. This is provided in Table 32. [Table 32]

A BASIC DIRECTION AND DIMENSION OF ROLL UP

Subsequently, a dimension used in the present invention will be described more briefly. Usually, time can be a dimension that can be a direction axis in roll up or drill down as various units such as date- week-month-quarter year-year and the like. Meanwhile, a patent classification code on a patent classification code system can be a dimension that can be a direction axis in roll up or drill down.

Firstly, the most basic time dimension can be selected from unit periods of one or more date-week-month-quarter year-year-a plurality of years, and the above multi-stage time dimension can be applied to many kinds of time attribute fields included in a patent document such as application date, publication date, registration date, etc.

Next, we refer to patent classification code dimensions. Since each country has introduced at least one patent classification code system, the patent classification code dimension can be established by reflecting multi-stage of a patent classification code system. Also, a bundle of at least one of the selected patent classification codes can be established as a separate dimension in one unit. For example, for a major theme of "RFID", a multi-stage minor theme corresponding to the major theme is established, and then an individual dimension can be established by corresponding a bundle of the patent classification codes to the minor theme. This individual dimension is particularly of use to a personalized multi-dimension analysis operation result table / (a personalized cube) a table / a cube made by an individual user to achieve the purpose of analysis.

Also, a region can be one dimension. A unit which can be easily obtainable from a patent document of region units is a nation, and a dimension having a multi-stage class can be generated by dividing from address information into regions in one nation.

A subject can be one dimension. The subject includes applicants, inventors, agents, etc., and the attribute of an applicant (the attributes of a corporation, a laboratory, a university, etc.), a quantitative attribute such as their scale (large, medium, small, etc.), an inter- applicant attribute such as head office-subsidiary and the like can also be a dimension. Also, corporation's financial information and corporation's monetary information such as a stock price, a sale, a profit ratio, etc. of corporations can be a dimension. Also, a meta attribute that can arbitrarily correspond to a corporation such as a global corporation, a local corporation, etc. can be a dimension.

Also, an attribute regarding conditions of a document such as application, registration, rejection, etc. can be a dimension. Also, if all kinds of counting values are grouped, a group each counting value belongs to can be a dimension. One example of these groups includes 1 through 5, 6 through 10, 10 through 15, and more than 15 claims. The number of joint applicants, the number of joint inventors, the number of family (the number of domestic family, the number of overseas family, etc.) and the like may not be a dimension in itself, but when the above number is grouped, a group the above number belongs to can be a dimension.

A multi-dimension analysis operation result table generating module 402 according to the present invention generates a multi-dimension analysis operation result table by performing a multi-dimension computation on at least one selected dimension. When the multi-dimension analysis operation result table generating module 402 according to the present invention generates the multi-dimension analysis operation result table, we suppose that a patent classification code is included in a dimension. Herein, a multi-dimension computation value for the given patent classification code is generated by considering the patent classification code and a higher patent classification code. When a multi- dimension computation value for any patent classification code is generated in this way and obtained, a value generated by reflecting the multi-dimension computed numerical value on all values for the patent classification code and its lower patent classification codes is obtained. When a given patent document includes a patent classification code, the multi-dimension analysis operation result table generating module 402 according to the present invention reflects a value resulting from performing a multi-dimension computation on the patent classification code on a higher patent classification code of the patent classification code as well as the patent classification code. For example, when document number #1 is assigned H04B 7/06 in IPC and multi-dimension computation data is generated by this document, H04B 7/06 is assigned the counting value of 1, and H04B 7/04 and H04B 7/02 that are immediate higher levels of H04B 7/06 should be assigned a counting value. Of course, it is obvious that higher levels more than H04B 7/00 are assigned a counting value.

As described above, a dimension can mainly be a direction axis of the roll up and drill down. Representatives of the dimension are a patent classification code such as IPC, etc. and time. Also, a nation, a region, an applicant, an inventor, status, citation, family information and the like can be a dimension axis. The multi-dimension analysis operation result table generating module 402 according to the present invention calculates an analysis index value beforehand using at least one of the selected dimensions as an object axis of roll up, wherein the analysis index value includes total amount, an occupying index, a concentration index, an activity index or the like.

Firstly, as seen from a space, a dimension can be subdivided into steps as entire nations - a separate nation - a region within a separate nation, such as entire nations - Korea - a special city of Seoul - GangNam district .... etc. (Herein, patent information includes a nation and/or address information of an applicant and/or an inventor). A dimension can also be divided in the same way, for example, time - degree - period —

If an applicant is subdivided, it will be subdivided as applicant

(entire nations) - applicant - (each nation). Also, an area of a middle scale (North America, Asia, Euroupe, Middle East, South America, Africa, etc.) can be introduced between the entire nations and each nation. To introduce the middle scale, the name of a nation corresponding to North America (for example, USA, Canada, Mexico or the like) needs to be mapped into a central category of North America. Further, roll up/drill down from an applicant to an inventor is possible, and examples of the direction of roll up/drill down will be provided below. 1) the roll up/drill down will be in the direction of applicant (entire nations), applicant, inventor of each nation. 2) the roll up/drill down will be in the direction of applicant (entire nations) applicant (entire nations) inventor of each nation. Meanwhile, the roll up/drill down will be according to the attributes of an applicant in one nation. For example, an applicant can be divided into organization units such as a corporation, a university, a laboratory, etc., the corporation can be divided into a large enterprise, enterprise of middle standing, and small and medium enterprises, the university can be divided into a national/public university, a private university, a college, etc., and the laboratory can be subdivided into a public laboratory and a private laboratory, and therefore, roll up/drill down regarding these things will be performed. Herein, mapping information for the applicant must exist. For example, if applicant "A" is a corporation, information that its scale is a small and medium enterprise needs to be mapped.

Also, if there is financial statements data of corporations or the like, more various roll up/drill down can be performed. For example, if there is information regarding sale, profit, stock price and rise and fall ratio of each of the formers as financial statements of a corporation, roll up/drill down in various categories such as corporation - sale scale - the range of the sale scale, corporation - profit - rise and fall ratio of the profit, or corporation - stock price - the range of the rise and fall ratio of the stock price can be performed. That is, the nature of the roll up/drill down means that if there are lots of category information on one object, the processing result according to each category is arranged. One patent document includes a nation, time, an applicant, inventors and patent classification code information, and the applicant information further includes region information of the applicant. Further, if an applicant is a corporation, financial statements, etc. of the corporation can be built as corporation information DB. Herein, 1) if one selected category such as patent classification code information in an entire or any defined patent document set unit has a multi-stage hierarchical structure, many kinds of information corresponding to patent classification codes selected in that category are processed (searched, counted or any other statistic and analysis calculations are performed) and the results can be output, and 2) roll up/drill down can be performed based on various intersections for more than two kinds of categories (also, it is obvious that when the kinds of the category is n, the possible number of the intersections are nCri (that is,n Combination r,and herein, r equals or is bigger than 1 and equals n or is smaller than n), all combinations may not be selected and some combinations can have a special analytic meaning.)in an entire or any defined patent document set unit.

In the case of roll up/drill down through the intersection, following points should be considered.

The first case is when using 2 categories. For example, when roll up/drill down is performed on category "A" (such as time) and "B" category (such as IFC), it is desirable to drill down in one direction (For example, a time axis as a detail IPC is not drilled down but only IPC is spreaded, which is shown when the concept of the roll up/drill down was explained in the above) for the benefit of expressing the results of information processing (search, counting or any other statistic and analysis calculations).

Of course, IPC is drilled down and a time axis can also be drilled down, but this can cause the result to be more complicated, so it is desirable to avoid it unless in a special case. Naturally, the roll up should be performed in both directions. That is, the roll up is performed based on the value of information for a cell comprising all combinations of a predefined unit of category"A"and a defined unit of category"B", and roll up information needs to be processed in a direction of category"A"in each cell and in a direction of"B"category. The immediate above table shows one example. Each number in a cell represents the value of information of the cell, a number in { } represents the number of patent registrations rolled up for each year from the year of 2000 to 2005 in each patent classification code level, and 17 in H04B 7/02 { } represents the number of patent documents rolled up for H04B 7/02 and up to its lower patent classification codes. The above table is rolled up with respect to years (for example, how many documents exist in the year of 2005, how many of them exist in H04B 7/02 and in 2005, etc. are indicated), but preferably they can be indicated after roll up, and it is preferable that multiple-dimension can be performed with respect to quarter years and months of 2005.)

The second case is when using more than 3 categories. Also, it is preferable that drill down is spreated in one direction, and spreading in two directions is allowed. However, it is preferred to avoid spreading in three directions because it is impossible or much difficult to express visually. Likewise, it is preferable that roll up has been computed for all categories, but it is better that roll up for a dimension axis which does not require beforehand roll up according to roll up computation amount and use frequency is not performed and then is processed when the use is started. That is, if roll up is not performed, necessary information can be rolled up and processed at the starting time, but it takes a relatively long time to process roll up. For example, if roll up for each year is not processed, the effect of drill down for each year occurs by obtaining year information from bibliographic data of all objective documents, dividing and calculating it according to each year, and outputting the result for each year. However, if information is divided and calculated beforehand, and rolled up, the rolled up year information just needs to be read and displayed later when drilling down for each year, thereby having an advantageous effect in the speed of a response.

Multi-dimension analysis operation result table data for total amount analysis

Hereinafter, a multi-dimension analysis operation result table that is generated by a multi-dimension analysis operation result table generating module 402 according to the present invention will be described in detail. Furthermore, the way the analysis module gains access to the multi-dimension analysis operation result table and what data the analysis module generates will be described. Typically, application date recitation mentioned in a patent document has the recitation of year-month-date. Therefore, if counting is performed in a daily unit and in each patent classification code, multiple-dimension computation can be performed in a monthly unit, in a quarter year unit and in a yearly unit. The following Table 33 shows a part of an embodiment of a table after multi-dimension computation is performed. Data provided in the following Table 33 refer to data that is generated by a multi-dimension analysis operation result table generating module 402, wherein the multi- dimension analysis operation result table generating module 402 generates the data by counting the number of a patent document beforehand in an applicant unit, in a yearly unit and in each layer of multi-stage IPC, based on an applicant, IPC and information on date for each year in a patent information table rebuilt with the Star Schema. Data such as Table 33 can be stored in any format such as table/view/materialized view or the like, and data included in a stored table can be extracted by using a proper query. For example, one example takes the number of applications for each year and for a multiple of application IPC (multiple-dimension computed) at the standard of IPC 1 dot subgroup of applicant "A" (hereinafter, referred to as "C5 level"). Also, it is preferable that the number of applicantions for applicants and years is generated. (To do this, it would be better to obtain data of the total amount of each year, (it is not indicated in the following Table 33 under the space restriction, but it can be obtained by addition for each year.)) In particular, the multi- dimension analysis operation result table generating module 402 needs to perform multi-dimension computation for information corresponding to lower patent classification codes of a patent classification code which is inaccessible by wild cards (*, ?, etc.) or the like with the reference to patent classification code DBs such as IPC and the like. (It is particularly important in a level under a subgroup in the case of IPC, in a level under a class level in the case of USPC and in a level where title information has a dot in the case of FT.) [Table 33]

In the following, Table 33 will be described. AppName refers to the name of an applicant, A number after C refers to node layers from a root node of IPC (Cl is a section, C2 is a class, C3 is a sub class, C4 is a main group, C5 is 1 dot sub group, C6 is 2 dot sub group, C7 is 3 dot sub group. Subsequently, C8 through C20 can be assigned. Typically, up to C15 will be enough), and 01, 02, etc. refer to years. Years exist before the year of 2000, and the number for this omitted. When the present time is assumed to be 2007, a numerical value based on application dates of documents published until now represents data in each cell. Meanwhile, when values are generated in the table, a value entered into each cell of a sum field represents an adding value of values in each year, and a gray period represents an adding value of values in each year belonging to a unit period from a year a date corresponding to one year and six months (normally, a laid-open publication period after an application date) before the present date belongs to until a present date.

GID is used to show the result that multi-dimension computation has been performed up to which level. H04B 7/02 of GID 7 shows the result that multiple-dimension computation up to itself (C5) has been performed

(for example, 64 = 24 + 39 + 0 + 1), and H04B 7/02 of GID 3, which is one step lower than that, shows the result that multiple-dimension computation up to C6 level (except itself) has been performed. 24 is obtained from GID 1, and 24 refers to the number of patent documents including a patent classification code of H04B 7/02. GID value can be recited in any way (this is arbitrarily defined.), but the present specification indicates it in a way of forming a series of "2^11""1" .

GID refers to the steps of roll up, wherein GID 0 indicates roll up to C8 level, GID 1 indicates roll up to C7 level, GID 3 indicates roll up to C6 level, GID 7 indicates roll up to C5 level, GID 15 indicates roll up to C4 level, GID 31 indicates roll up to C3 level, and GID 63 indicates roll up to C2 level. Herein, it is obvious that GID can be assigned to levels before C8 level and Cl level, but this is ineffective and thus applying it to system 1 does not matter. In H04B 7/06, GID 0 means roll up to C8 level and GID 1 means roll up to C7 level. (The above example does not have C8, and therefore a value of roll up to C7 equals that of roll up to CS.)

An embodiment of the mentioned GID and its relationship with levels of patent classification codes will be described with reference to Table 34 below. [Table 34]

GID Objcet scope IPC level USPC level FT level ECLA level

The GID indicates a code that means roll up to the corresponding level. For example, GID 15 indicates C4 level. This means that information based on roll up to a main group in IPC, to 0 dot in USPC, to 00 level in FT, to a main group in ECLA (roll up of itself and up to its all lower patent classification) is written in that line. Therefore, if information in any level is wanted to be seen, multi-dimension computed information up to that level can be seen by specifying GID corresponding to that level. In this specification, GID is a short term, and it means "roll up level code". In the above and in the following, the roll up level code and GID are used as a same meaning. In FT, 00 level means a level that consists of two alphabetic letters without any numeric digit such as AA, etc. A "supercategory" indicates that USPC or FT has lots of items in a class and a theme code level, respectively, and therefore the items have been grouped. As can be seen from Table 33, each patent classification code in each line has its immediate higher patent classification code on the left. The multi-dimension analysis operation result table generating module according to the present invention generates multi-dimension computed numerical value per each IPC symbol as data provided in Table 33. This is generated with respect to all patent classification codes included in each patent document group of all patent document groups of applicant "A" in each nation. In addition, data provided in Table 33 can be generated regarding all application documents, or it can be generated regarding only all registered documents. The multi-dimension analysis operation result table generating module 402 generates a mutl-dimension analysis operation result table such as Table 33 for each document set having a predefined attribute. The multi-dimension analysis operation result table generating module 402 includes the steps of 1) obtaining a key value (document unique information) characterizing a patent document, such as the number of the patent document included in a patent group having the predefined attribute, 2) extracting key values characterizing the patent document from FACT table of a table rebuilt with the Star Skima, and 3) generating a multi- dimension analysis operation result table regarding at least one predefined analysis theme with respect to only patent document corresponding to the extracted key value. After performing the steps of 1) and 2), the multi-dimension analysis operation result table generating module 402 generates to-be-processed patent document set which will be subject to processing. The FACT table includes FACT information for entire document sets. Obviously, the FACT information includes document unique information, and further includes many kinds of bibliographic data information. Desirably, the document unique information includes application number or document unique code or document series code.

Desirably, the multi-dimension analysis operation result table generating module 402 generates at least one multi-dimension analysis operation result table regarding at least one predefined theme with respect to all documents in a predefined analysis theme, in each nation or in entire nations. The document set can be any document set which shares at least one predefinable attribute. Examples of these any document set includes 1) a patent document set which is generated by a method of specifying a patent document set with a specific patent classification code on a specific patent classification code system in a specific nation DB, 2) a patent document set which is generated by a method of specifying a patent document set with the name of a specific applicant in a specific nation DB, 3) a patent document set which is generated by a method of specifying a patent document set with the name of a specific inventor included in a patent document of a specific applicant as an inventor (That is, including the name of an applicant and the name of an inventor) in a specific nation DB, 4) a patent document set which is generated by a method of specifying a patent document set with the name of a specific agent in a specific nation DB, 5) a patent document set which is generated by a method of specifying a patent document set with the name of a specific applicant and a specific patent classification code on a specific patent classification code system in a specific nation DB, 6) a patent document set which is generated by a method of specifying a patent document with the name of a specific applicant, the name of a specific inventor and a specific patent classification code on a specific patent classification code system in a specific nation DB, 7) a patent document set which is generated by a method of specifying a patent document with the name of a specific applicant and the name of a specific agent in a specific nation DB, 8) an entire patent document set of a specific nation, 9) all patent document sets of at least two nations, and 10) a patent document set specified by a period predefined in the steps of 1) through 9). Herein, the above any document set may be generated just with respect to family information regarding other nations besides a first nation. Also, in the case of USA, the above any document set may be a document set consisting of ony reissued patent documents. The document set mentioned in this paragraph is exemplary. Therefore, for these exemplary document sets, the multi-dimension analysis operation result table generating module 402 desirably generates at least one multi-dimension analysis operation result table regarding at least one predefined theme with respect to the above documents according to a predefined analysis theme.

The multi-dimension analysis operation result table generating module 402 can generate a multi-dimension analysis operation result table such as Table 33 regarding any document set generated by a user. The multi-dimension analysis operation result table generating module 402 includes the steps of 1) obtaining a key value (document unique information) characterizing a patent document, such as the number of the patent document included in any patent group generated by a user, 2) extracting key values characterizing the patent document from FACT table of a table rebuilt with the Star Skima, and 3) generating a multi- dimension analysis operation result table regarding at least one predefined analysis theme with respect to only patent document corresponding to the extracted key value.

A processing method of a total amount data multi-dimension analysis operation result table generating module 402 To generate a multi-dimension analysis operation result table such as Table 33, the multi-dimension analysis operation result table generating module 402 performs the following steps on the predefined/obtained document set (For example, all document sets applied for or registered in a specific nation).

Firstly, a key value (document unique information) characterizing a patent document, such as the number of the patent document included in a patent group having the predefined attribute is obtained, only key values characterizing the patent document from FACT table of a table rebuilt with the Star Skima are extracted, and then a command set for at least one predefined analysis theme-related multi-dimension analysis basic table is obtained regarding only the patent document corresponding to the extracted key value. (This command set can be script.) (Herein, an analysis theme- related multi-dimension computation execution command set with respect to the multi-dimension computation basic table can be obtained simulataneously). In the case of an entire patent document set comprised of all documents applied for or registered in a specific nation, an analysis theme-related command set to generate at least one predefined analysis theme-related multi-dimension operation result table is obtained without any other procedures. That is, in this step, to-be-processed patent document set on which the multi-dimension analysis operation result table generating module 402 perfoms a predefined processing is determined and a command set for each analysis theme to perform the predefined processing is determined. Secondly, the multi-dimension computation basic table is generated by executing the command set for each analysis theme with respect to a separate document of the to-be-processed patent document set. The multi- dimension computation basic table includes information on at least one dimension, and the each record needs to have basic data for each dimension to perform multi-dimension computation. The basic data include information on existence/non-existence (1 or null) (for example, application/registration has occurred or has not occurred) and many kinds of counting information (more than one information mentioned as counting information in this specification, for example, the number of claims, etc.). The dimension includes more than one selected from the following dimensions or more than one of a complex dimension, that is, a combination of the following dimensions. In particular, the dimensions include 1) a patent classification code dimension for the kinds of each patent classification code, 2) a date dimension, 3) a position dimension (a nation, a region, etc.), 4) a subject's attribute dimension (a dimension distinguishable in the scale, the attribute, etc. of more than one selected from an applicant, an inventor or an agent (for example, the applicant is divided as a corporation, a university, a laboratory, an individual and the others, the corporation is divided as a multinational corporation, a large enterprise, small and medium enterprises, etc.) can be obtained from subject master DB 204 with reference to the attribute of each applicant.).

Herein, a field must be provided in a mult i-dimension computation basic table to input bibliographic data needed for each dimension. A patent classification code dimension of the field includes a field for each level in order to accommodate all higher or lower patent classification codes on a patent classification code system such as from a section to n dot subgroup with IPC, and (from a supercategory) from a class to n dot level with USPC, etc. Also, it would be better that a field in the smallest unit for analysis exists for a time dimension. In the case of patent analysis, an analysis up to a monthly field will be enough when a daily or weekly unit analysis is not needed. Also, if an applicant attribute dimension exists, a field corresponding to a corporation (a multinational corporation, a large enterprise, small and ^'medium enterprises), a university, a laboratory, an individual and the others is needed.

As one example, if including a patent classification code dimension and an application date dimension regarding an applicant in the multi-dimension computation basic table are sufficient, data provided in the following Table 35 regarding a document that applicant "A" applied in Jan. 3, 2005 and IPC is H04B 7/06 only have to be generated. [Table 35]

Also, a period and applicant attribute may be input like the following. Data provided in the following Table 36 resulted from performing a partial mult i-dimension computation in a year or an applicant attribute, which can be processed at the time of multi-dimension computation, or can be generated beforehand. [Table 36]

Also, the position of an applicant, corporation evaluation such as a financial level of an applicant, a family number dimension with respect to a category of the number of family, a claim number dimension with respect to a category of the number of claims and the like can be included. As one example, the following Table 37 has additional data below. [Table 37]

In the following, a method of generating such data is provided. 1) a patent classification code included in a document is queried of patent classification code master DB 203, all patent classification codes of the patent classification code is obtained, and all higher patent classification codes obtained are input according to each level. For example, if IPC is H04B 7/06, H04B 7/06 is input to 3 dot subgroup (C7) level, H04B 7/04 is input to 2 dot subgroup (C6) level, H04B 7/02 is input to 1 dot subgroup (C5) level, H04B 7/00 is input to main group (C4) level, H04B is input to subclass (C3) level, H04 is input to class (C2), and H is input to section (Cl) level. 2) A value needed for a time dimension is input with respect to a necessary standard date (for example, an application date) of many kinds of date information such as an application date, a registration date, etc. If an application date is a standard date (for example, Jan. 3, 2005) and time dimension is divided as a monthly dimension, a quarter year dimension and a yearly dimension, Jan. 1, 2005 is input. 3) other bibliographic data are input to other dimension. Then, 1 is input to a large corporation. A field value such as the position of an applicant is filled by using address information, the number of family of a patent document is queried by inquiring of the patent document master DB 202, which category a queried value belongs to is determined, and 1 is recorded in that category. A range of the number of claims is processed like the number of family.

If one patent document has more than two kinds of patent classification codes (for example, IPC and USPC), the multi-dimension computation basic table of each kind can be generated, or more than two kinds of patent classification codes can be processed within one table. Also, if a document includes a multiple of patent classification codes of the same kind, it is desirable to generate an individual record with respect to each patent classification code. Also, if an applicant is more than two joint applicants, generating records for each applicant which have the same information for each applicant is desired. (If a subject is an inventor, a record is generated for each inventor, and if a subject is an agent in the case of agent analysis, a record for each agent is generated.) Thirdly, multiple-dimension computation with respect to the multi- dimension computation basic table is performed, and a multi-dimension analysis operation result table is generated. Table data provided in the following Table 38 shows the result of performing roll up computation regarding up to a yearly dimension of time dimension, with respect to IPC dimension. (For convenient recitation, roll up computation regarding documents of the year of 2004 in a quarter year-year unit was indicated). [Table 38]

The method of performing multiple-dimension computation is divided into a method of performing roll up computation and a method of performing cube computation. Hereinafter, the roll up computation and the cube computation, respectively, will be described with respect to a dimension such as IPC, a date and an applicant. When IPC dimension, date dimension and applicant dimension exist, the roll up computation proceeds in one direction selected from 3 dimensions, but the cube computation proceeds in 3P3 direction (the number of permutations occuring when 3 of 3 are selected, herein, 6). That is, when IPC dimension is 1, date dimension is 2 and applicant dimension is 3, the roll up computation proceeds in one direction such as l->2->3, but the cube computation proceeds in all six directions including l->3->2, 2->l->3, 2->3->l, 3->l->2, 3->2->l as well as l->2->3, or proceeds in one or more direction selected from the above directions. Therefore, if a computation result up to l->2->3 (Finally, a roll up computation result in an applicant unit occurs) is not needed while the roll up computation is performed, that is, if a roll up computation result in only IPC is needed or if a roll up computation result in only IPC and date dimension like Table 38 is needed, a roll up computation is performed by designating a desired roll up direction only with respect to one or more selected dimensions.

The multi-dimension analysis result table generating module 402 obtains application date information and patent classification code information from a given patent document, and extracts higher patent classification codes of the obtained patent classification code with reference to patent classification code master DB 203, which includes data of a patent classification code system or a table rebuilt with Star Skima. Data of the multi-dimension analysis operation result table including information such as the above table is completed using the extracted higher patent classification codes and year information of the application date. Herein, one patent document includes more than two kinds of patent classification codes (for example, USA has IPC and USPC), each kind of patent classification code is independently processed. Furthermore, if one patent document includes two or more patent classification codes of one kind, 1) only the first main patent classification code can be processed, 2) all patent classification codes can be processed, 3) and if main patent classification codes and sub patent classification codes have weights, they can be processed by reflecting the weights. In the case of 3), if a numerical value has a decimal point, rounding off the number will be a proper inscription method. Herein, which policy to take out of the above 1), 2) and 3) is optional. Also, another policy besides 1), 2) and 3) can also be selected. When the method 2) is taken, a line number of the multi- dimension analysis operation result table increases since at least one new data value is generated for each patent classification code (that is, roll up computation for sub IPC is needed). Also, when the method of 3) is taken, the numerical value can include a decimal point.

The table is an index regarding a quantity such as the number of applications or the number of registrations of patent indices. Data processing such as roll up/drill down including lower patent classification codes according to the present invention can be used for many kinds of indices such as an occupying index, a concentration index and an activity index.

GENERATING A MULTI-DIMENSION ANALYSIS OPERATTION RESULT TABLE WITHOUT AN ANALYSIS THEME-RELATED MULTI-DIMENSION COMPUTATION BASIC TABLE

We described that the multi-dimension analysis operation result table is generated using an analysis theme-related multi-dimension computation basic table. Herein, the analysis theme-related multi- dimension computation basic table is not necessary, and if a table rebuilt with Star Skima structure exists, the multi-dimension analysis result table generating module 402 according to the present invention can generate the multi-dimension analysis operation result table. It is because the analysis theme-related multi-dimension computation basic table can reduce the calculation amount/the information processing amount of the multi-dimension analysis result table generating module 402, the one analysis theme-related multi-dimension computation basic table can be used in generating multiple tables. The multi-dimension analysis result table generating module 402 can perform the multi-dimension computation by obtaining necessary information from a table rebuilt with Star Skima structure and processing it if the analysis theme-related multi-dimension computation basic table does not exist.

GENERATING A MULTI-DIMENSION ANALYSIS OPERATION RESULT TABLE WITHOUT A TABLE REBUILT WITH STAR SKIMA STRUCTURE

The multi-dimension analysis result table generating module 402 can read necessary data from not a table rebuilt with Star Skima structure, but the patent document master DB 202, the patent classification code DB 203 and/or the subject master DB 204 and generate the analysis theme- related multi-dimension computation basic table. In this case, a program (script or the like) generating the analysis theme-related multi-dimension computation basic table can be relatively less complicate, and recycling capability of the entire program or each module forming that program can be relatively lowered. For example, the analysis theme-related multi- dimension computation basic table stores information on IPC and all its higher IPC for each application number, wherein the multi-dimension analysis result table generating module 402 according to the present invention finds IPC included in the application number, queries of the patent classification code master DB 203 or all other higher patent classification code table, etc., obtains information on all higher IPCs of the found IPC and generates the analysis theme-related multi-dimension computation basic table based on this.

GENERATING AN ANALYSIS THEME-RELATED MULTI-DIMENSION ANALYSIS OPERATION RESULT TABLE FROM PATENT DOCUMENT MASTER DB 202

To generate the multi-dimension analysis operation result table, the multi-dimension analysis operation result table generating module 402 performs the following steps regarding the predefined/obtained document set (for example, all document sets applied for or registered in a specific nation).

Firstly, a key value (document unique information) characterizing a patent document such as the application number, etc. of the patent document included in a document set having a predefined attribute is obtained and stored. Also, in order to generate an analysis theme-related multi-dimension operation result table such as the following Table 38, at least one data of material data of Table 35 through Table 37 for each patent document is obtained and generated from the patent document master DB 202, the patent classification code master DB 203 and/or the subject master DB 204, the generated data is combined, and the combined data is stored in a memory. The data stored in a memory can be the same data as the data stored in the multi-dimension computation basic table. That is, data stored in the memory includes information on at least one dimension, and the each record has to include each dimension basic data for performing multi-dimension computation. The basic data includes information on existence/non-existence (1 or null) (for example, an application/a registration has occurred or has not occurred) and many kinds of counting information (more than one information mentioned as counting information in this specification, for example, the number of claims, etc.). The dimension includes more than one selected from the following dimensions or more than one of a complex dimension, that is, a combination of the following dimensions. In particular, the dimensions include 1) a patent classification code dimension for the kinds of each patent classification code, 2) a date dimension, 3) a position dimension (a nation, a region, etc.), 4) a subject's attribute dimension (a dimension distinguishable in the scale, the attribute, etc. of more than one selected from an applicant, an inventor or an agent (for example, the applicant is divided as a corporation, a university, a laboratory, an individual and the others, the corporation is divided as a multinational corporation, a large enterprise, small and medium enterprises, etc.) can be obtained from subject master DB 204 with reference to the attribute of each applicant.).

Secondly, the multi-dimension operation result table generating module 402 performs predefined analysis theme-related multi-dimension computation on data stored in the memory by combining at least one information of Table 35 through Table 44 for each patent document and generates an analysis theme-related multi-dimension analysis operation. IPC LEVEL RANKING INFORMATION FOR AN INDIVIDUAL APPLICANT When a multi-dimension analysis operation result table such as Table 38 is given, the analysis module can generate the following analysis result.

Firstly, when an applicant is obtained, the number of applications/the number of registrations-based ranking information for each level of patent classification codes is generated with respect to the obtained applicant. The ranking information is generated by comparing count values of the multi-dimension computed application or registration document with respect to each level of at least one patent classification codes of the obtained applicant. For example, a numerous-filed/numerous- registered ranking for each patent classification code in 1 dot subgroup level of applicant "A" can be obtained by comparing a counting value of the number of multi-dimension computed documents in H04B 7/02 of 1 dot subgroup level (C5 level) of applicant "A" and a counting value of the number of mult i-dimension computed documents in another patent classification code of 1 dot subgroup level of applicant "A" from among the multi-dimension analysis operation result table data.

The following Table 39 provides a multi-dimension analysis operation result table which has counted the number of applications (or registrations) for each applicant, each IPC level and each year. [Table 39]

The multi-dimension analysis operation result table has been generated with respect to each applicant of all applicants in a specific nation. For example, when a patent classification code and an application date included in a specific document of applicant "C" are provided, a counting value of a specific cell (a patent classification code, an applicant (AppName)) corresponding to the above Table is created with reference to a patent classification code system and by using the patent classification code and the application date data of the patent document or a number can increase. If a multi-dimension operation result table for all applicants in a specific nation unit such as the above Table is generated, the analysis module can generate the following analysis result from the multi-dimension analysis operation result table.

Firstly to show IPC subclass level grouped information of applicant "A" , a yearly counting value in this level can be generated. Preferably, a value of a recent field is a counting value based on patent documents obtained from a corresponding year that one year and six months (18 months) prior to a predefined basis date (one example thereof includes an inquiry date of a multi-dimension analysis operation result table) belongs to until the inquiry date of a multi-dimension analysis operation result table. (... refers to the corresponding value.) The following Table 40 shows one example. [Table 40]

Of course, when H04B of Table 40 is drilled down, a counting value for documents corresponding to a patent classification code that is an immediate lower code of H04B in a patent document of applicant "A" such as the following Table 41 is generated. This comes out by counting on condition that "AppName = A and IFC level = C4" in the multi-dimension analysis computation result table such as the following table, and then it is ordered by IPC level = C4 (herein, C4 refers to an IPC subclass).

When drill down is again performed in Table 41, a multi-dimension computed counting value comes out with respect to 1 dot subgroup as described above. (Of course, when H04B is drilled down, counting values for documents corresponding to patent classification codes which are immediate lower codes of a patent classification code of H04B in patent documents of applicant "A" come out as indicated below.) This is generated by counting on condition that "AppName = A and IPC level = C5" in the multi-dimension analysis operation result table such as the above table, and then it is added and ordered by IPC level = C5 (herein, C5 refers to IPC main group). A counting value in each cell is a value resulting from performing a multi-dimension computation on all values corresponding to its lower patent classification codes. When H04B 1/00(+) is drilled down, its lower patent classification code comes out, a value in each cell comes out by counting on condition that "AppName = A and IPC level = C6", and then it is ordered by IPC level = C6 (herein, C6 refers to IPC 1 dot group). A counting value in each cell is a value resulting from performing a multi-dimension computation on all values corresponding to its lower patent classification codes. Drill down can be performed continuously until its lower patent classification code does not exist any longer, and values needed in cells can be generated by entering the above condition until an additional drill down may not be performed. Such method has been described regarding IPC, but it is obvious that this also can be applied to USPC, FT, FI and ECLA.

In the above paragraphs, a multi-dimension computed multi- dimension analysis operation result table generating method, a cell value calculating method and a calculated value indicating method during drill down have been described with respect to applications (or the total amount of registrations), but it is well known in the art that the raulti- dimension analysis operation result table can be generated with respect to an occupying index, a concentration index and/or an activity index.

*Also, a multi-dimension computed value with respect to other indices corresponding to a predefined definition equation which includes a patent classification code in such a way can be generated in the same method. Also, a cell value during drill down can be calculated, and the calculated value can be indicated.

When the above multi-dimension analysis operation result table is provided and at least one of specific patent classification codes are given, a ranking of applicants and yearly counting values of the applicants can be generated in higher count value order. For example, If H04B is given, adding and ordering for each applicant on condition that "IPC level = C4 and C4 = H04B" are performed with respect to each applicant (that is, ordered by AppName) . ONE EXAMPLE OF GENERATING AN OCCUPYING INDEX AND OTHER DATA

For example, if an occupying index is defined as "a value equal to dividing the number of the documents of an object document set having specific attributes by the number of the documents of an entire document set having specific attributes, data similar to the above Table 39 can be generated with respect to the occupying index. That is, if an occupying index with respect to a specific patent code of a specific corporation "A" is obtained, an object document set will be a document set including a specific patent classification code of applicant "A" document set, and an entire document set will be an entire document set including a specific patent classification code. Herein, it is obvious that when a document set including the patent classification code is obtained, documents including the patent classification code and its lower patent classification code should be included.

Also, the multi-dimension analysis operation result table generating module 402 according to the present invention can store a computation result regarding a patent index which can be defined by other equations as well as an occupying index using the same method as the total amount analysis.

Hereinafter, analysis indexes will be introduced. 1. A technical innovation activity concentration degree analysis indexes include a) revealed technological advantage, 2) revealed Patent Advantage, 3) concentration Ratio n (CRn), 4) herfindahl Index (HHI), etc.

2. A technical level analysis index includes 1) patent count weighted by citations, 2)) cites per patent (CPP), 3) patent impact index

(PII), 4) current impact index (CII), 5) technology strength (TS), 6) technology cycle time (TCT), 7) science linkage (SL), 8) average claims per patent, and 9) family size, Etc.

3. A cooperation relation and knowledge flow analysis index includes 1) number of patents with co-applicants, and with co-inventors),

2) salton's index, 3) brain gain, brain drain, and 4) index for knowledge flow with patent citations.

It is possible to additionally introduce analysis indexes having higher applicability than patent indexes. All these analysis indexes can be expressed as a formula. It will be obvious to those skilled in the art that all formulas can correspond to a query formula. Reference ranking generation method

A reference ranking generation method will be described in detail. When the acquirement of applicant A is determined, the document set of applicant A can be obtained (determined) from the first nation or the second nation, IPC, etc. can be extracted from the document set, the number of corresponding documents of IPC in each step can be counted in the extracted IPC according to each step (all from a section to n dot subgroup), the total number of numerous applications/numerous registrations, concentration, the ranking of an activity field can be calculated by using the counted value according to each IPC level (ranking information generation according to each patent classification code level of patent index reference applicant A has been described above). The high ranking field in can be extracted by the calculated ranking, and a patent classification code of at least one extracted IPC, etc. can be a high ranking technical field. When one subject (or a characteristic if it is expanded) is given, the above described scheme becomes a scheme for extracting the ranking of the poorest patent classification code of each step in at least one corresponding patent classification code about the given subject (for example, respective to an inventor, agent, etc, ranking according to each patent classification code level can be generated based on a document set including own name of the inventor, agent, etc. by the same scheme).

FIG. 19 shows a view, in which the reference ranking is generated according to each level of the patent classification code system, such as IPC, etc., according to each nation, such as Korea, U.S, Japan, Europe, etc, based on application or registration and according to each application respective to the number of applications/registrations, an occupancy ratio, and a concentration ratio and/or activity ratio. In a case of Samsung Electronics of FIG. 19, HOlL is in the first rank, and H04N is in the second rank in an IPC sub-class level base on Korea application documents.

This will be described below in detail. It is possible to extract the ranking of the poorest patent classification code according to each level of at least one corresponding patent classification code in the first nation or the second nation through the steps described below, the steps including 1) determining at least one document set having a common characteristic (for example, in a case where at least one common characteristic is included, such as common application A, common inventor A or agent B of applicant A, or a specific common keyword is included, there is a common specific time period, citing a document of a document set having a specific characteristic or the document is cited, etc., a partial document set, which includes the common characteristic, of the entire document set can be generated through an inquiry by means of SQL or a search engine), 2) extracting at least one corresponding patent classification code respective to each document constituting a document set, 3) obtaining all higher patent classification codes of the extracted patent classification code based on the patent classification code system, 4) storing all the higher patent classification codes according to each patent classification code levle (for example, in a case of IPC, from a section to n dot sub-group), 5) counting each document including the stored patent classification code through a roll-up calculation according to each level, 6) performing calculation according to each predetermined analysis index with reference to a counting result (the step of performing calculation according to each predetermined analysis index, such as the total amount calculation, concentration ratio calculation, activity ratio calculation, etc.), and 7) calculating ranking in consideration of roll-up according to each patent classification code level based on each analysis index (a rank command among SQL commands may be used). At this time, it is natural that the document set may have a specific common characteristic (for example, an applicant characteristic, such as applicant A (it may be a plurality of applicants, as used herein), period limitation, such as the 5 recent years, etc., inventor limitation, such as inventor C, and/or a characteristic obtained by combining respective characteristics (inventor C of applicant A, etc.)). It is obvious that the above described steps 1) to 7) can be calculated according to each applicant included in the first nation in a case of a document set related to each applicant.

The multi-dimensional analysis operation result table generating module 402 of the present invention can generate multi-dimensional analysis operation result table data by performing at least one of steps 1) to 7).

Give attention to the fact that ranking according to each numerously-filed IPC of the IPC sub-class level has been generated in the left column of the view in FIG. 19. Numerical values shown in the present view showing the implemented embodiment are obtained by processing values, which have been input into the analysis module, by using a computational method, and the numerical values are variable according to supplementation, deletion, and alternation of data, etc. Therefore, it is possible that the numerical values are different from actual values based on a specific time point. Therefore, it is obvious that the sprit of the present invention doesn't depend on numerical values shown in various tables, etc. included in the specification of the present invention, but depends on construction, configuration, a tool, and a method for generating the numerical values, an information processing procedure, a system, an application method, etc. Hereinafter, the above described concept is the same. FIG. 19 is a view showing an embodiment of an analysis result generated when drill-down is continuously performed respective to HOlL. When performing drill-down, an analysis information value only about an object, to which drill-down has been performed, is generated. At this time, if AJAX technology is used, it is possible to generate only a numerical value respective to a part where change (drill-down) rapidly has occurred without reloading the entire page and provide the numerical value to the user. In the present invention, the user has to be interpreted as a user computer 300 in view of relation between the user and the system of the present invention. That is, in view of the system, 1) "something is provided to the user"actual Iy means "something is transmitted to a user's computer," 2)"the user inputs something" actually means "a value input by the user is received by the user's computer," and designation, specification, and selection by the user, etc. actually mean receiving information, which has been designated, pre-specified, and selected by the user, from the user's computer. Therefore, in the specification of the present invention, it is appropriate to interpret expressions relating to the user as described above. However, the reason why expressions based on the user are shown is to make descriptions convenient/more easily understandable.

FIG. 19 is a view showing an embodiment of an analysis result generated when drill-down is performed to HOlL 21/00 and other lower patent classification code. It can be understood that an analysis result based on the total amount is obtained through drill-down. The drill-down can be performed down to the lowest level of IPC, and it is possible to perform drill-down down to the lowest level as long as data exists therein. Also, this is equally applied to other patent classification codes.

FIG. 31 is a view showing an example of a technical field analysis utilizing the patent classification code of the present invention, and is a view showing an embodiment of an application total amount analysis including drill-down respective to IPC H04B based on entire U.S application documents and lower classification thereof. At this time, all pieces of data corresponding to a lower patent classification code, which have been rolled up, are included in the analysis data of each cell. Citation analysis-multi-dimensional analysis operation result table

In order to perform a citation analysis, the multi-dimensional analysis operation result table generating module 402 generates a citation analysis multi-dimensional analysis operation result table data. A procedure where the multi-dimensional analysis operation result table generating module 402 generates a citation analysis multi-dimensional analysis operation result table data will be described with reference to U.S patent application No. 09/802,847 (the applicant: Samsung Electronics Co., LTD(KR), the title- user request processing method and apparatus using upstream channel in interactive multimedia contents service, and application data: 2001.3.12).

The bibliographical information of U.S patent application No. 09/802,847 is shown in Table 42. [Table 42]

Herein, the multi-dimensional analysis operation result table generating module 402 obtains citation parent-child data as described below from the bibliographical information of patent application No. 09/802,847 in relation to citation. The child refers to the document number of a corresponding document itself, and the patent refers to the document number of other document cited by the corresponding document (at this time, the document number may also be an application number, and may also be a registration number). Particularly, although it is preferred that the document number is a number, which allows the corresponding document to be specified, it is desirable that the document numbers are unified into application numbers, which are common numbers that all documents have, because it is possible that a publication number or application number exists in the parent data. However, it has also no problem to unify document numbers into registration numbers. Also, if the complexity of processing information doesn't become a problem, and there is no problem in specification of documents, it is also possible to use various numbers together with each other (in this specification of the present invention, for convenience in description, document numbers are unified into application numbers). Data shown in Table 43 refers to obtained data reference-based citation parent-child data. [Table 43]

As shown in Table 43, parent-child data is generated based on a unified application number from the obtained citation parent-child data as shown in Table. 44. Each application number is obtained from the registration number and the publication number (there is no problem if it is an application number, all pieces of number information have each document type, and A registration, Al publication, etc.) of the parent (an application number or an application serial number is used in U.S, but this is only a title. Therefore, characteristics of data indicated by the title are the same. In the above and below description, an application number is used as a unified number). Data of the table below refers to a unified number reference-based citation parent-child data. Meanwhile, nation columns of documents in child and parent columns show source nations, to which a corresponding document relates (is applied or registered). Such source nations may be various nations as well as U.S. Although the nation columns are omitted hereinafter in order to achieve convenience in description, it has to be understood that nation columns exist. Meanwhile, the notation form of the application number may be an arbitrary notation, such as "double digits + / + six digits" or "year +6 digits, " etc. [Table 44]

Although the data processing shown in Table 44 is performed based on one given document, such data processing can be performed according to at least one given document set and according to each document included in the document set. A result obtained through such data processing is shown in Table 45. At this time, it is possible to allow the repetition of parent (application number) according to each document set (i.e. when document 1 and document 2, which are included in one document set, include document 3 in citation information, duplicates of the document 3 can be allowed). That is, it is possible to generate unified number reference- based citation parent-child data respective to all individual documents included in a document set. At this time, the most representative document set may be a US open application document set and a US registered patent document set. It is obvious that a document set generated by the user (which is output as a next search result after inputting a specific search formula) and a document set automatically generated by a computation system 1 performing the present invention can be the most representative document set. [Table 45]

In addition, if searching for a patent document master DB 202 respective to all documents included in the unified number reference-based citation parent-child data, data, to which bibliographical items are added according to each document may be generated as shown in Table 46. At least one selected among all elements included in the bibliographical items may be added. Table 46 below shows data, to which an application date and a registration date among the bibliographical items corresponding to respective application numbers are added one by one. [Table 46]

It is also possible to add count information of each application number (various pieces of count information, such as the number of claims, the number of drawings, the number of families, the number of inventors, the number of applicants, etc.) to the data show in Table 46 according to each count field. It is desirable to generate one table respective to an item corresponding to each application number one by one.

Subsequently, the above described data can be generated as a field value, in which at least two pieces of data can correspond to each application number. These may be an applicant, inventor, agent, patent classification code, etc. The patent classification code will be described later. In a case where at least two pieces of data correspond to each application number, all necessary field values can be input into one table. However, if it is, a problem of increase in the size of the table has to be considered. For example, applicants of patent application No. 09/309,895 shown in Table 47 are AT¹T Corp. (New York, NY) and Sun Micro Systems (Palo Alto, CA). In this case, as shown in table 48 below, an application field has to be added to the same row of Table 47 below relating to the application number. [Table 47]

[Table 48]

As such, if there are two applicants, the number of lines is doubled. Meanwhile, if there are six inventors respective to the application, 12 lines are required (2*6=12) in a case where an applicant and the inventors are shown in one table. At this time, there is a problem in that repetition occurs in all pieces of data except for data relating to a pair (an applicant and an inventor) (of course, in a case where repetition occurs, there is no table join when an analyzed result (e.g. an applicant of each year and an inventor of each year, who is a specific applicant) is obtained, so that the increase in the size of the table becomes a problem).

For reference, it is preferable that data processing based on time/date is performed as described below in consideration of roll-up. For example, in a case of March 12, 2001, it is preferable to perform data processing after departmentalizing the data into March 12, 2001, March, 2001, and the first quarter of 2001, for roll-up calculation. However, in the specification of the present invention, due to a notation problem (that various kinds of numerous fields are expressed on a screen with limited width) departmentalized data are omitted. However, it is obvious to those skilled in the art. Meanwhile, in relation to patent classification code, the data processing will be described in view of lower classification code inclusion. First, as shown in Table 49 below, data processing about application number, 09/802,847, corresponding to a child will be described, and the same is applied to a document included in the parent. In 09/802,847, IPC codes of G06F 15/16 and H04N 7/16 exist, and USPC codes of 709/203, 709/217, 709/231, and 725/135 exist. [Table 49]

In Table 49, C3 refers to an IPC subclass level, and it has been described above that others follow the same organization. The numeral 1 recorded in the field of 2001 means that there is one document in 2001, and the numeral 1 recorded in the field of March, 2001 means that there is one document in a corresponding month based on each month. The above mentioned notation method is a typical notation method used for data construction in consideration of roll-up. Although both G06F 15/16 and H04N 7/16 have been included in 1 dot subgroup, if H04N 7/169 is included in the document with the application number of 09/802,847, Table 49 above would be the same as Table 50 below (this is arbitrarily introduced for describing the spirit of the present invention). [Table 50]

A result obtained by processing the data respective to USPC in view of lower classification code inclusion is shown in Table 51 below. [Table 51]

USPC 709/203 (title information: client/server) exists in 709/201 (title information: distributed data processing), which is directly higher, and is included in class 709. The same is applied to other codes. Although all of the USP, etc have been included in ldot or no dot level, if 725/45 is included in the document with application number of 09/802,847, Table 51 above would be the same as Table 51 below (this is arbitrarily introduced for describing the spirit of the present invention). [Table 52]

H04L 12/56, H04L 12/28, USPC7} 370/395, and 370/235 of IPC have been allocated (the IPC may be different from IPC given by Korea patent office, but data is processed according to each nation so that U.S data is preferably processed by IPC given by U.S patent office). At this time, one row is generated by each patent classification code. Therefore, although IPC code and USPC code may be generated in one table, it is preferred that they are not generated in one table. Particularly, in a case where multiple codes, such as multiple applicants and multiple inventors, exist, it is further preferred that the codes are not generated in one table. Essentially, this is because one child corresponds to several parents. That is, one chide data may correspond to parent data, the number of pieces of the parent data being K, one chide data may correspond to IPC codes, the number of the IPC codes being m, and USPC codes, the number of the USPC codes being n, and each IPC code and each USPC code may exist in respective parent documents, the number of the parent documents being K. Therefore, if they are gathered into one table, a significantly large number of lines is required. At this time, multiple pieces of data are repeated. This repetition occurs more often in a case where the size of a document set is large (e.g. a U.S registration patent entire set, etc.). Therefore, it is preferable to separate IPC codes and USPC codes from each other. In relation to a patent classification code, there may be table types, such as 1) a table type where only patent classification codes of the child side of citation child-parent data have been processed, 2) a table type where only patent classification codes of the parent side of citation child-parent data have been processed, 3) a table type where only the patent classification codes of the both child and parent sides of citation child-parent data have been processed. In the case of table type 3), there may be a problem of increase in the number of rows, but there is an advantage in that possibility of table join is reduced.

That is, the multi-dimensional analysis operation result table generating module 402 generates unit citation parent-child data according to each document included in a given/preset document set and obtains at least one of bibliographical items of each document included in the child and/or parent, respective to each piece of unit citation parent-child data, thereby generating multi-dimensional analysis operation result table data having the obtained bibliographical item as field content.

*As an example of the given/preset document set, there may be 1) a specific nation-application document or registration document entire set, 2) a document set based on at least one of specific IPC/USPC codes, 3) a document set according to each specific applicant, 4) a search-based generation document set, which can be generated by a specific search formula, etc., 5) an entire nation integration-all application document or all registration document set, etc.

At this time, if data shown in Table 52 is generated respective to 1) an entire document set according to each nation, it is possible to obtain usefulness described below. It is natural that bibliographical information corresponds to each application number/registration number of the data, and at least one piece of the bibliographical information is generated through a combination as described above. [Table 53]

A method for generating such data is a method for inputting all application/registration document numbers into a child column and inputting numbers of parent documents citing the application/registration documents into a parent column respective to each application/registration document. When there is no parent value, a null value is input. At this time, data exist in only child column, and a parent document respective to a corresponding child document does not exist. In this case, all document numbers is input into the child column.

The simplest model is introduced for convenience in description. It is assumed that each of document 1, document 2, document 3, and document 4 includes one piece of citation information, and a citation order is in the order of document 1 -> document 2 -> document 3 -> document 4 (it means that document 2 cites document 1, document 3 cites document 2, and document 4 cites document 3). At this time, based on document 2 (document 2 exists in the child column), document 1 exists in the parent column of document 2, and based on document 3 (document 3 exists in the child column), document 2 exists in the parent column. Also, based on document 4, document 3 exists in the parent column. That is, all documents 1 to 4 exist in the child column, and at least, documents 1 to 3 exist in the parent column. In consideration of citation depth (based on document 4, document 1 has citation depth 3, document 2 has citation depth 2, and document 3 has citation depth 1), document 1 and document 2 exist in the range to citation depth 2 of backward citation (a document set generated by citation information included in document 3) relating to document 3, and document 4 exists with forward citation depth 1. In this case, since documents 1 to 4 exist in the child column, it is possible to find a citation document with citation dept 1 in the parent column based on any document, and when the found citation document number is again found in the child column, and a parent column corresponding to the found child column is found, a citation document with citation depth 2 can be found. Such a model, will be described below with Table 54. [Table 54]

Although one document is used in the above-description, it is obvious that the same may be applied to a document set. An application to a document set will be described by using a model shown in Table 55. Based on document set 3, this will be described (in Table 55 below, it has to be understood that each cell (the content within a line) isn't, in actual fact, cell data of a multi-dimensional analysis operation result table, but is one set consisting of cell data of several lines in view of conception, and this understanding has been done for convenience of describing the spirit of the present invention). [Table 55]

In Table 55 above, all document sets 1 to 4 become a subset of an entire set of the child column, and all document sets 1 to 3 become a subset of an entire set of the parent column. The sentence"the document sets become a subset" means that all documents included in the subset are included in the entire set so that the documents can be checked in and extracted from the entire set. A characteristic of a document set may be an arbitrary characteristic and also may be a characteristic of 1) an applicant name, 2) inventor name, 3) IPC or USPC of each step, 4) nation, 5) agent, 6) term range (an application date/registration date range), 7) document state (under an application procedure, registration, rejection, etc.), or 8) combination of at least one among the above-mentioned items as a common item.

For example, it is assumed that the document set of all registered documents, which are included in IPC HOlL, the applicant of which is Samsung Electronics, and have been registered in U.S, is document set 3, document set 2 consisting of all citation documents with backward citation depth 1 relating to document set 3 may be extracted. Meanwhile, it is possible to obtain document set 4 consisting of all cited documents with forward citation dept 1 relating to document set 3 and document set 1 with backward citation depth 2 relating to document set 3. At this time, document set 1, document set 2, and document set 4 can be analyzed. 1) the total amount of backward/forward citation, 2) a backward/forward numerously-citing applicant, 3) a backward/forward numerously-citing inventor, 4) backward/forward numerously-citing IPC/USPC (drill-down according to each step in each patent classification code system is the same as described above), and 5) each backward/forward numerously-citing document can be found as an object respective to the analysis. Meanwhile, in the case of 1), it is possible to display the total amount after classifying it according to each-year's/preset period, and it is also possible to know an increase/decrease ratio and an increase/decrease rate based on the various kinds of numbers. In the case of 2), it is possible to classify and display backward/forward citation amount of each nemerously-citing applicant according to the total amount and/or each year's/preset period. The same is applied to cases of 3) to 5). When a forward nemerously-citing applicant of 2) is selected as an object to be analyzed based on document set 3, it is possible to find information about a later-filed applicant citing a U.S registration document included in IPC HOlL of Samsung electronics. Samsung electronics may be included in such a later-filed applicant set. In this case, this becomes self citation. If a numerical value of the self citation is divided by other numerical value or is pre-processed, various analyses respective to self citation are possible. Further, when later-filed applicants are considered according to each rank, it is possible to check an applicant citing many documents included in IPC HOlL of Samsung Electronics. Of course, a link, which allows the user to obtain a related document, is shown on the analyzed result (the number of forward citations). When the link is clicked by the user, a related document (a forward citation document) is obtained from the patent document master DB 202 to be provided to the user. Meanwhile, when one document set (for example, document set 3) has been given, it is natural that document set 2, which is the parent of document set 3, is found, document set 2"' satisfying a condition preset in document set 2 (e.g. a term condition, such as a document applied within the 10 recent years, etc.) is generated, at least one preset analysis is performed after document set 2" ' is set as a to-be-analyzed patent document set, and an analyzed result can be generated.

4 types of to-be-analyzed document set 4 types In the meantime, regarding the citation analysis, 4 types of to- be-analyzed document set will be available. The following 4 types of to- be-analyzed document set are determined for a single reference document set relating the citation analysis.

FIG. 38 illustrates 4 types of to-be-analyzed document set relating to the citation analysis with respect to the reference document set. Referring to FIG. 38, 4 types of to-be-analyzed document set include

1) "other company's patent cited by own company' s patent (self→other)" ,

2) "other company' s patent citing own company' s patent (other→self)" ,

3) "own company' s patent cited by other company' s patent (other→self)" , and 4) "own company' s patent citing other company' s patent (self-→other)" .

Type 1) is an entire backward citation document set cited by an individual reference document included in the reference document set. The entire backward citation document set can be determined as union of every document corresponding to the citing information (reference cited) included in the respective reference documents (at this time, the duplication may be removed, but the duplication may be allowed in the union). For example, if document #1 and document #2 include document a as the citing information, document a is cited two times so that it may considered as an important document in the view of the counting, and thus it is preferred that the duplication is allowed (an identical record is duplicative in the parent in the multi-dimensional analysis operation result table. Two records relating to parent a are present resulted from documents #1 and #2). In the meantime, it is preferred to remove the duplication (through union operation) simply in document a aspect (applicant, date, other bibliographical particulars). In the analysis of the present invention, it is more reasonable to allow the duplication because the analysis is performed on the counting.

Type 2 is an entire forward citation document set citing the individual reference document included in the reference document set. The method of obtaining the entire forward citation document set includes 1) a method of searching the document number citing the individual reference document and summing the searched document number and 2) a method of searching for the reference document number in the patent column of the child-parent table generated for the entire patent document, searching for at least one or more child document number corresponding to the reference document number included in the parent column in the child column, and summing the searched numbers. Method 2) is more preferable.

Type 3) is an entire forward citation-relating reference document set. The entire forward citation-relating reference document set is a document set including every reference document that had been cited by another document from among every reference document included in the reference document set. The entire forward citation-relating reference document set includes the parent document of the document set of type 2). Type 4) is an entire backward citation-relating reference document set. The entire backward citation-relating reference document set includes the reference document including citing information from every reference document included in the reference document set. The entire backward citation-relating reference document set includes the child patent document of the document set of type 1).

In FIG. 39, a topmost tap includes 1) citation analysis for the patent document entire set, 2) technology area-based citation thorough analysis, 3) inventor-based citation analysis, 4) multi-citation individual document-based citation analysis, and those are all related to the determination of the reference document set. In 1) citation analysis for the patent document entire set, the patent document entire set serves as a reference document set determined by any one selected from all the document determining methods relating to the determination of the document set of the present invention. In 2) technology area-based citation thorough analysis, the document set departmentalized from the patent document entire set as the reference document set in which the patent document entire set is departmentalized according to the patent classification code that is organized according to the patent classification code (IPC, USPC, etc.) level-based patent classification code reference ranking serves as the reference document set. In 3) inventor-based citation analysis, the document set obtained through extracting an inventor form the patent document entire set, calculating the ranking, and departmentalizing according to the inventor having the higher ranking serves as the reference document set. In 4) multi-citation individual document-based citation analysis, the individual document obtained through researching the number of citing/cited information of every document included in the patent document entire set and extracting, the document having the higher citing/cited ranking serves as the reference document set. Multi-dimensional analysis operation result table generating module 402

*The multi-dimensional analysis operation result table generating module 402, if the reference document set is determined, determines 1) the entire backward citation document set, 2) the entire forward citation document set, 3) the entire backward citation-relating reference document set, and/or 4) the entire forward citation-relating reference document set as the to-be-citation-analyzed document set corresponding to the determined reference document set, generates the multi-dimensional analysis operation result table described in the present specification with respect to any one document set of 1) to 4), and stores the generated table.

Result of citation analysis

The analysis module of the present invention obtains the analysis result of the analysis subject-based multi-dimensional analysis operation result table according to the determined 1) the entire backward citation document set, 2) the entire forward citation document set, 3) the entire backward citation-relating reference document set, and/or 4) the entire forward citation-relating reference document set with a preset analysis formula. In the meantime, the analysis module can conduct various analysis described in the present invention for any one or more of document sets 1) to 4), not for the multi-dimensional analysis operation result table. However, this case has a problem in that it is impossible to obtain the multi-dimensional operation result or it requires much computation resources for obtaining the multi-dimensional operation result, so as not to be much recommended.

The analysis module accesses the multi-dimensional analysis operation result table generated according to document sets 1) to 4) and extracts the object citation analysis result with the preset analysis formula. The object citation analysis result includes the followings.

First, the analysis module extracts, according to the applicant information on, Oapplicant ranking, ii) ranking in the unit of at least one level-based patent classification code in at least one patent classification code system according to the applicant, iii) multi-patent inventors ranking, and iv) the document frequency reference ranking, numerical value data of the respective rankings calculated by roll-up operation according to the time dimension or other dimension, and/or the document number corresponding to the respective rankings. That is, the analysis module can generate information on the document number for obtaining the specific individual document itself corresponding to the ranking value according to the various fields included in the bibliographical particulars, the value calculated by the roll-up operation according to the time or other dimension, and/or the various values with respect to to-be-citation-analyzed document sets 1) to 4). Second, the analysis module extracts, centering on the patent classification code, information on i) the ranking in the unit of at least one level-based patent classification code in at least one patent classification code system, ii) applicant belonging to the level-based patent classification code, iii) the ranking of the inventors belonging to the level-based patent classification code, and iv) the query document frequency reference ranking belonging to the level-based patent classification code, numerical value data of the respective rankings calculated by the roll-up operation according to the time dimension or other dimension, and/or the document number corresponding to the respective rankings. That is, the analysis module can generate information on the document number for obtaining the specific individual document itself corresponding to the ranking value according to the various fields included in the bibliographical particulars_', the value of the ranking calculated by the roll-up operation according to the time or other dimension, and/or the various values with respect to to-be-citation- analyzed document sets 1) to 4).

Third, the analysis module extracts, centering on the applicant attribute (a kind of applicant (company, university, individual, etc.), in the case of the company, financial attribute, company evaluation index, etc.), information on i) a kind of applicants, ii) the company size evaluation index, such as the sales, iii) the ranking according to the company financial evaluation elements, such as an year-average profit ratio, and/or the numerical value data of i) to iii) calculated by the roll-up operation according to the time dimension or other dimension, and/or the document number corresponding to i) to iii). That is, the analysis module can generate information on the document number for obtaining the specific individual document itself corresponding to the ranking value according to the various fields included in the bibliographical particulars, the value calculated by the roll-up operation according to the time or other dimension, and/or the various values with respect to the to-be-citation-analyzed document sets 1) to 4).

Fourth, the analysis module extracts the data of the variation value, such as increase/decrease rate and increase/decrease speed of the numerical value data if the numerical value data is present in above first to third information.

The (based on the applicant citation) to-be-citation analyzed multi-dimensional analysis operation result table generated by the multi^¬ dimensional analysis operation result table generating module 402 is mainly generated subjecting to the patent document cited in the applicant's own document.

In the meantime, it is obvious that an identical multi-dimensional analysis operation result table can be generated subjecting to the examiner-cited document. That is, if considering the document cited by the examiner in the examination for the specific application document (child document) as the parent document of the application document, all the actions of the multi-dimensional analysis operation result table generating module 402, its resultant (with respect to the examiner citation) to-be-citation analyzed multi-dimensional analysis operation result table, and a method of using the resultant are completely identical to the case with respect to the applicant's citation. In the meantime, for the combined citation analysis, it is possible to generate the combined citation document set (parent-combined) through union operation of the applicant-cited document (parent) and examiner-cited document (parent 2) with respect to a single application document. Further, if the above processing with respect to the applicant-cited document is identically performed to the combined citation document set, it is possible to obtain an identical resultant and the method for using the identical resultant is identical. If referring to the description with respect to the applicant- cited document for understanding the description of this paragraph, it is obvious to those skilled in the art so that the repetitive description will be omitted.

Hereinafter, it will be described with reference to the drawings in more detai1. FIG. 38 illustrates a year-based analysis result of citation total amount for the to-be-analyzed document set according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set. FIG. 38 illustrates that a total amount tap is selected. The accompanied drawings show various taps in the analysis image according to the exemplary embodiment, in which if the character of the tap is thick, it means that the tap is selected. A kind of taps basically includes 1) the reference for the attribute of the document set when determining the document set, such as the application document reference or registration document reference, 2) the reference of the application/registration total amount, occupancy rate, concentration rate, activity rate, or other analysis index, 3) the reference relating to the nation determination, such as Korea, US, Japan, and Europe, 4) the reference relating to an object or a subject of the analysis, such as the total amount, applicant, inventor, technology, and individual document, 5) the reference for various numerical values, increase/decrease rate of the numerical value, and increase/decrease speed of the numerical value, 6) the reference for selection of the patent classification code, such as IPC, USPC, FT, FI, and ECLA and the level of the patent classification code.

FIG. 39 illustrates a year-based analysis result for the multiple- citing applicant according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set.

FIG. 40 illustrates a year-based analysis result for the multi- citing patent classification code (IPC main group level) according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set. At this time, the IPCs can be drilled down in which the citation analysis result for the sub-patent classification code relating to the drill down is extracted to be provided to the user, which is illustrated in FIG.41.

FIG. 41 illustrates a year-based analysis result reflecting the drill down of the multi-citing patent classification code (IPC main group level) according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set. FIG. 42 illustrates a year-based analysis result of the multi- citing inventor according to an exemplary embodiment of the present invention in which, when all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set. FIG. 43 illustrates a year-based analysis result of the most citing applicant with respect to the to-be-analyzed document set according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set .

FIG. 44 illustrates a year-based analysis result for the most citing inventor according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set.

FIG. 45 illustrates a year-based analysis result when the IPC in the analysis for the most citing technology according to the IPC main group is drilled down according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set.

FIG. 46 illustrates a year-based analysis result when the IPC in the analysis for the most citing cited technology of the USPC sub class (no dot, class direct subordinate) is drilled down according to an exemplary embodiment of the present invention in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set .

FIG. 47 illustrates a total amount reference analysis result and a chart of the analysis result generated by a chart generating module 406-2 among a reporting module 406 of the present invention for the total amount reference analysis result with respect to the to-be-analyzed document set in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the entire forward citation document set serves as the to-be-analyzed document set. The reporting module 406 of the present invention includes at least one of a table generating module 406-1 for generating a table, a chart generating module 406-2 for generating a chart, a graph generating module 406-3 for generating a graph, and a report generating module 406-4 for generating a report .

FIG. 48 illustrates the analysis result of a year-based citation total amount of the often-cited inventor and a document list generated by a simplified analysis module 407 that provides the simplified analysis result (document list, the number of applications/registrations of the most-filing applicant according to the year, the year-based number of applications/registrations according to the most-filing inventor, the number of applications/registrations of the most-filing technical field (IPC, USPC, FT) according to the year (including drill-down) for the document set corresponding to the specific number when the specific number represented in the analysis result is clicked, with respect to the to-be- analyzed document set in which, when all the US patent registrations in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the forward citation document set serves as the to-be-analyzed document set.

FIG. 49 illustrates that the simplified analysis module 407 provides a function of the drill-down according to the most-filing technology field (IPC, USPC, FT). FIG. 50 illustrates a year-based analysis result of the citation total amount with respect to the to-be-analyzed document set, in which, when the numerously-filed IPC sub class unit document in all the US patent applications in the name of the Samsung Electronics among all the applicants in the DB of the patent information system 1 are considered as the reference document set, the backward citation document set serves as the to-be-analyzed document set .

Competition analysis multi-dimensional analysis operation result table The multi-dimensional analysis operation result table generating module 402 for conducting the competition analysis generates competition analysis multi-dimensional analysis operation result table data. A method of generating the competition analysis multi-dimensional analysis operation result table data by the multi-dimensional analysis operation result table generating module 402 will be described.

Competition in view of the applicant

The competition in view of the applicant can be defined as the conflict between applicant A and other applicants B belonging applicant A's nation (a first nation) or at least one another nation (a second nation). In the meantime, the conflict between the applicants may occur when they involve commonly in 1) the numerously-filed technical field, 2) high-concentrated field, and 3) high-activated field, and it is preferred that the technical field in the numerously-filed/high-concentrated/high- activated technical field is defined with at least one patent classification code according to the level, such as IPC according to the level. For example, applicant A of a first nation files several applications in HOlL based on the IPC subclass, the competitor includes 1) an applicant having the multiple applications in HOlL (an applicant having high occupancy rate), 2) an applicant having high concentration rate in HOlL, and 3) an applicant having high activity rate in HOlL, on a basis of the IPC subclass in a first nation (in a second nation, 1), 2), and 3) applicants can be direct or potential competitors). In the meantime, applicant A highly concentrates and activates in HOlL, 1), 2), and 3) applicants of a first nation or a second nation can be the competitors.

Method of obtaining applicant-view competition information

Hereinafter, the method of obtaining applicant-view competition information will be described. In the above, the method of generating the ranking information according to the level of the patent classification code based on the respective analysis indexes (total amount, occupancy rate, concentration rate, activity rate, etc.) in the respective nations of applicant A has been described. For example, when H04B 7/02 in the IPC

1 dot subclass (C5 level) on a basis of the Korean patent application document in the name of the Samsung Electronics is selected as the numerously-filed subclass IPC, the multi-dimensional analysis operation result table generating module 402 generates the following information.

The multi-dimensional analysis operation result table generating module 402 extracts 1) a numerously-filed/numerously-registered applicant, 2) an applicant highly concentrating on H04B 7/02, 3) a highly active applicant, or 4) an applicant having a high calculation value according to other patent analysis indexes with respect to the application document/registration document in a first nation. The extracting method can be basically processed with the SQL query. At this time, if the multi- dimensional analysis operation result table data has been generated according to the level of the patent classification code based on the total amount, occupancy rate, concentration rate, and activity rate, the mult i-dimensional analysis operation result table generating module 402 accesses the multi-dimensional analysis operation result table data with a simple SQL query so as to obtain at least one applicant information highly related to the target competition and the amount of applicant's application/registration in the unit of the preset time period. If the multi-dimensional analysis operation result table data has not been generated, the target information can be obtained by using a relatively long and complex SQL query having the following steps.

The logic of the SQL query includes a command processing the steps of 1) extracting every document including the obtained specific-level patent classification code (e.g. IPC H04B 7/02) and the sub-patent classification codes of the specific patent classification code (the sub- patent classification codes are searched in the patent classification code master DB 203 and obtained, which is described previously.) from the patent document master DB 202 in the unit of the nation/nation unification (in this case, if an identical document is present two times or more, the duplication must be removed.), 2) obtaining the information on the applicant and date, such as application date/registration date, from the bibliographical information of the extracted document, 3) sorting the obtained information based on the numerously-filed/numerously-registered applicant and obtaining ranking information on the numerously- filed/numerously-registered applicant, and/or 4) counting the amount of applications/registrations in the unit of the preset term from the application date/registration date information. It is of course that any two or more steps 1) to 4) can be processed at one time. The SQL logic is described based on the total amount, but can be processed by a similar manner with respect to other patent analysis index, such as the occupancy rate, concentration rate, and activity rate.

For example, in order to search the highly concentrating applicant, even after generating the applicant list after step 2), the number of applications/registrations for H04B 7/02 of the respective applicants is divided with the total number of applications/registrations of the respective applicants (which can be obtained from the patent document master DB 202 with the SQL) so as to calculate the concentration rate, and the sorting of step 3) can be processed with the calculated concentration rate. If the mult i-dimensional analysis operation result table generating module 402 has generated the multi-dimensional analysis operation result table data according to the level of the patent classification code, the processing can be processed with the simple SQL query (such the simple SQL query is very easy to the person skilled in the art).

If the sub-patent classification code of the patent classification code in step 1) cannot be easily obtained with the extension (*, ?, etc.), there is a problem in that the data processing may be burdened with much loads in step 1). Therefore, in this case, the data as shown in Table 55 generated by the multi-dimensional analysis operation result table generating module 402 based on the counting information on the document relating to the sub-patent classification code through the roll-up operation is more useful. [Table 56]

If there is given the data of the multi-dimensional analysis operation result table as shown in Table 56, the record including H04B 7/02 in the IPC 1 dot level (C5 column) is extracted and the number of records is grouped-by with the applicant field of AppName so as to grade the ranking. Through such the process, it is possible to extract competitive applicant B, etc. based on H04B 7/02 of applicant A and to provide the number of applications/registration according to the year and term of applicant B to the users through reading the processed value of the documents including H04B 7/02 and its sub-patent classification code from the documents of applicant B generated in the multi-dimensional analysis operation result table.

The method of generating at least one mult i-dimensional analysis operation result table data for competition analysis by the multi^¬ dimensional analysis operation result table generating module 402 includes the following steps. First, the multi-dimensional analysis operation result table generating module 402 generates the mult i-dimensional analysis operation result table data as shown in Table 56 according to the preset type of competition. When the mult i-dimensional analysis operation result table generating module 402 generates the table/data as shown in Table 65 with the occupancy rate, concentration rate, activity rate, or other analysis index, the AppNarae, the level of the patent classification code, etc. are identical and the numerical value according to the year/term becomes the value of the occupancy rate, concentration rate, activity rate, or other analysis index, respectively.

In the above, it has been described the method of extracting 1) a numerously-filed/numerously-registered applicant, 2) an applicant highly concentrating on H04B 7/02, 3) a highly active applicant, or 4) an applicant having a high calculation value according to other patent index based on the application document/registration document for the obtained IPC H04B 7/02 in a first nation, but 1)~4) can be extracted by a same manner for the obtained IPC H04B 7/02 in a second nation (the obtained H04B 7/02 can be identically processed by a same manner above based on the patent document data of a second nation. Of course, if the respective nations include the unique patent classification code, the patent classification codes have the multi-stage hierarchical structure like the IPC and the hierarchical structure can be classified by the number of the dots so that the patent classification codes can be identically processed as in the IPC.)

FIG. 33 illustrates a representative competing company analysis based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system 1 of the present invention according to the exemplary embodiment of the present invention.

The analysis of the representative competing company according to the applicant is subjecting to the patent classification code of the IPC group level among the patent classification code of the numerously-filed applicant obtains the ranking and the amount of the applications of the patent classification codes, grading the score by the preset calculation formula based on application-amount information of the applicant who files multiple applications for the patent classification code, and provides the ranking together with the applicant having the high score. It is preferred that the representative competing company analysis also shows information on the number of applications/registrations of the representative competing companies according to the year. Of course, if the number of applications/registrations is clicked, the document corresponding to the number of applications/registrations is obtained with the query formula included in the number of applications/registrations, to be transmitted to the simplified analysis module 407, and the simplified analysis module 407 provides the simply analyzed information on the documents. The numerical value included in every cell of the present invention is based on the query value and a basis document generating the corresponding numerical value from the respective query formulas can be obtained, which is applied to every cell. If the numerical value is a ratio value, the query corresponds to the respective numerical values according to the indexes included in the formula generating the ratio so that it is possible to obtain a basis document based on the query.

FIG. 33 illustrates a competing applicant analysis according to the numerously-filed patent technology classification code based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system 1 of the present invention according to the exemplary embodiment of the present invention. The obtaining the competing applicant according to the level of the specific technology classification code and the technology classification code can 1) obtain ranking information (HOlL of first ranking, H04N of second ranking in FIG. 33) on the reference patent classification codes according to the specific technology classification code level, 2) extract the applicant filing a lot of applications (based on the application amount), having high occupancy rate (based on the occupancy rate), having high concentration rate (based on the concentration rate), having high activity rate (based on the activity rate), or having a high value of other patent index is extracted together with the ranking in the specific nation (which is selected by a user or automatically by the system 1 among Korea, US, Japan, Europe, etc. according to the patent classification based on the specific technology classification code level (e.g. HOlL in FIG. 33)), and 3) provide the number of applications/registrations of the applicant together with the extracted applicant information. At this time, the code is drilled down, if above 1) to 3) are performed for the sub-patent classification code of the drilled-down patent classification code including even one document of the applicant so as to provide information on the competing applicant. Up to the lowermost patent classification code can be drilled down. In the meantime, the analysis of the competing company can be conducted by the same method with respect to USPC, FT, etc., as well as the IPC, but there may be a limitation on the nation selection (because the USPC is the US patent classification code so that it is impossible to obtain the competing applicant with the Korean and Japanese patent document). The competing applicant can be generated based on not only the application total amount of the applicant, but also the registration total amount of the applicant.

In the meantime, the reference patent classification code is generated based on the patent document of the specific applicant of a first nation and the numerously-filed, highly concentrating, highly occupying, and highly active competing applicant can be extracted in a second nation, not in a first nation, according to the reference patent classification code. A first nation and a second nation can be selected by the user or by the default value of the system 1.

In the meantime, it is possible to analyze an entering competing applicant and the analysis information for the entering competing company can be generated only with respect to the document within a preset recent term (e.g. for recent 7 years). That is, 1) based on the document of the applicant filed/registered within a recent term in a first nation, the information on the reference of the patent classification code according to the patent classification code level is generated, and the multi- application having, highly concentrating, highly occupying, and highly active competing applicant according to the reference of the patent classification code on a basis of the i) recent term or ii) entire term can be extracted in a first nation or a second nation. 2) Based on the document of the applicant filed/registered during an entire term in a first nation, the information on the reference of the patent classification code according to the patent classification code level is generated, and the multi-application, highly concentrating, highly occupying, and highly active competing applicant on a basis of the i) recent term can be extracted in a first nation or a second nation according to the reference of the patent classification code. Competition in view of the inventor

In the above, the method of generating the data by the multi^¬ dimensional analysis operation result table generating module 402 for obtaining the competition information in the applicant view has been described. The processing of the competition in view of the inventor (generating the document set based on the inventor, generating the table/data based on the inventor-based document set by the same manner, and accessing the data on a basis of the inventor) can be generated by a same method as that of processing the competition in view of the applicant (generating, in the unit of the document set unit based on the applicant, the table/data is generated according to the analysis index, such as the total amount, occupancy rate, concentration rate, and activity rate, based on the patent classification code level, and accessing the data on a basis of the applicant), an example of which is represented in Table 57 below.

Inventor analysis multi-dimensional analysis operation result table

There are two types of the data generated by the multi-dimensional analysis operation result table generating module 402 is generated on a basis of the inventor.

First, the data is generated through placing the inventor under a single applicant and processing the document set based on the inventor of the single applicant (i.e. when it is assumed that the applicant is generally a large enterprise or organization and the inventor belongs to the enterprise or organization, the inventor is linked with the applicant under. The data can be generated according to the document set satisfying the condition, such as applicant A AND inventor 1, applicant A AND inventor 2,...). Second, the data is generated by processing the document set based on the inventor only with the inventor independently (the inventor is inputted in the place of the applicant). An example relating to the former is obvious so that the exemplary embodiment of the data scheme generated by the multi-dimensional analysis operation result table generating module 402 for the latter is represented in Table 57 below. [Table 57]

The method of generating the data, such as the occupancy rate, concentration rate, activity rate in the unit of the individual inventor of the specific applicant by the multi-dimensional analysis operation result table generating module 402 when the total information is given is identical to that of generating the information based on the applicant. Of course, if the data of Table 57 includes information on every inventor of all the applicants, the multi-dimensional analysis operation result table generating module 402 may generate the data of the value of the patent analysis index, such as the occupancy rate, concentration rate, activity rate, on a basis of the total amount information in the unit of the individual inventor, not the inventor under the specific applicant. FIG. 35 illustrates a numerously-filed inventor-based analysis result according to the year based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system 1 of the present invention according to the exemplary embodiment of the present invention. The analysis of the inventor according to the patent classification code level-based area includes the steps of 1) generating the ranking of the patent classification code reference according to the patent classification code level of the specific applicant, 2) executing any one of i) calculating information on the most application inventor according to the year and ii) calculating information on the ranking of the numerously-filed inventor according to the year unification (e.g. from 1^st to 10^th ranking) according to the ranking of the patent classification code reference, and 3) providing the inventor analysis information extracted through the execution to the user. In the meantime, the analysis of the inventor according to the patent classification code level-based area does not limit the applicant in the unit of the nation, but can be conducted on a basis of the application document or registration document. At this time, the analysis of the inventor on a basis of the application document or registration document includes the steps of 1) generating the ranking of the patent classification code reference according to the patent classification code level based on the entire subject document, 2) executing any one of i) calculating information on the most inventor according to the year and ii) calculating information on the ranking of the numerously-filed inventor according to the year unification (e.g. from 1^st to 10^th ranking) according to the ranking of the patent classification code reference, and 3) providing the inventor analysis information extracted through the execution to the user.

FIG. 36 illustrates a numerously-filed inventor-based analysis result according to the year based on the total amount of the Samsung Electronics among all the applicants in Korea in the DB of the patent information system 1 of the present invention according to the exemplary embodiment of the present invention. The generation of the information on the inventor individual-based analysis for every inventor (the inventor belonging to the applicant of the company) relating to the specific applicant in the unit of the nation includes the steps of 1) obtaining every application document or registration document relating to the applicant in the unit of the specific nation, 2) extracting the inventor from the obtained document, 3) counting the number of documents according to the inventor based on the preset policy, 4) generating the ranking according to the inventor, and 5) generating the information according to the most application inventor and the number of applications/registrations based on the year/specific term according to the ranking so as to provide the generated information to the user. In the meantime, if the inventor is not related to the specific applicant, every patent document in the unit of the nation is obtained and steps 2) to 5) are executed. The documents can include only the documents belonging to the preset term, such as recent 7 years. This is an essential issue of generating the document set and it is okay that the document set can be generated by any one from the methods of generating the document set variously described in the present specification.

Patent technology classification-based technology analysis multi^¬ dimensional analysis operation result table

*973Next, the patent technology classification utilization technology analysis will be described. The patent technology classification utilization technology analysis means the analysis of the document set determined through the patent classification code among the document sets. A kind of the patent technology classification utilization technology analysis document set generated according to the patent classification code includes 1) a document set generated with the document including at least one patent classification code selected from the kinds (IPC, USPC, FT, ECLA, etc.) of the specific patent classification codes (if two or more patent classification codes are involved, various operations, such as OR, AND, NOT operations, between the patent classification codes are included.) and 2) a document set (IPC AND USPC, etc.) including two or more kinds of patent classification code. Both 1) and 2) document sets can be 1) in the unit of the individual nation or 2) in the unit of the nation union, and the duplication in the unit of the nation must be removed (a result of the union operation). Further, 3) whether to process the document set including only the main patent classification code or including every patent classification code is determined according to the preset policy or the user's selection, which are all related to the determination of the document set.

The information obtained from the patent technology classification utilization technology analysis (hereinafter, referred to as the technology analysis) is the information on the 1) occupancy rate, 2) concentration rate, and 3) activity rate, and 4) other patent analysis index of the patent technology classification utilization technology analysis document set (hereinafter, referred to as the technology analysis document set). At this time, it is of course that the documents corresponding to the sub-patent classification code of the patent classification code given in the data processing of above 1) to 4) are included in the to-be-analyzed document set.

The total amount analysis is first described from the patent technology classification utilization technology analysis. The total amount analysis is to provide the analysis result in the amount view, such as the number of term-based/year based applications/registrations according to the patent technology classification code level, in which up to the document including the sub-patent classification code of the specific patent classification code is calculated by the rolled-up operation so that the multi-dimensional operated result is stored (the result is a numerical value calculated by summing up the number of patent documents including the specific patent classification code and the number of patent documents including the sub-patent classification code of the specific patent classification code(it is of course that the duplication is removed.)). Table 58 represents the multi-dimensional analysis operation result table for a year-based distribution of the total application amount of the USPC according to the exemplary embodiment. USPC 002048000 is equal to 2/48 and USPC 002049100 is equal to 2/49.1, which are dependent on the marking methods (if three places in a front side referring to a class number and the number obtained by dividing 6 places in a rear side by 1000 are placed at a right and left side of "/", respectively, this corresponds to the USPC typically indicated in the pub1ication document . ) [Table 58]

The number of applications/registrations in the obtained specific USPC can be identified through obtaining the multi-dimensional operated number or total according to the year/term in the specific USPC level (class to n dot) column.

In the meantime, the multi-dimensional analysis operation result table generating module 402 can generate the multi-dimensional analysis operation result table data as shown in Table 59 for the applicant-based multi-stage patent classification code-based analysis. [Table 59]

In the data as shown in Table 59, when the multi-applicant is extracted on a basis of C12N 15/00, the result of the roll-up operation up to the document count information relating to C12N 15/00 (an IPC sub-group is shortly referred to as C4 level.) itself and its sub-patent classification code based on the IPC C4 level is GID 15, "GID = 15 and IPC = C12N 15/00 and IPC level = C4" is commanded in the condition formula in the above table, the number of applicant-based documents satisfying the condition (the number of documents represented in GID 15 row) comes out. When the total amount counting information is given, as described previously, it is possible to generate the data, such as the multi- dimensional analysis operation result table according to the analysis index, such as occupancy rate, concentration rate, and activity rate, but the multi-dimensional analysis operation result table as shown in below is suggested for the exemplary embodiment. In the meantime, the fact that the equal operation can be performed with respect to other patent index, such as the occupancy rate, is obvious to one skilled in the art.

Convergence analysis mult i-dimensional analysis operation result table

Hereinafter, the multi-dimensional analysis operation result table data generated by the multi-dimensional analysis operation result table generating module 402 for the convergence analysis will be described.

The multi-dimensional analysis operation result table generating module 402 of the present invention generates the multi-dimensional analysis operation result table data below for the convergence analysis of the present invention. The multi-dimensional analysis operation result table generating module 402 receives the inputted, established, or determined patent document set. When the multi-patent classification codes of one type are combined in the document set (in only IPC, two or more IPCs are processed, in the IPC and USPC, if there are two IPCs, they are is processed, but only one IPC is not processed, and in two or more USPCs, they are processed with the USPC multi-dimensional analysis operation result table). At least one type of patent classification code, such as IPC, USPC, FI, FT, ECLA, is included in a single patent document and the patent classification code includes at least one information. It shows that the data as shown in Table 72 includes at least one essential main patent classification code and selective sub-patent classification code on a basis of the document number (here, the type of the main patent classification code is identical to that of the sub-patent classification code.) At this time, it can be identified that the patent classification code included in the document arrays the patent classification code itself and its every higher patent classification code. When assuming that the main patent classification code of the patent classification code itself is corresponding to IPC 2 dot sub-class, the multi-dimensional analysis operation result table generating module 402 determines the main patent classification code as C6 level referring to the patent classification code master DB 203, inputs the main patent classification code into C6 level, inputs the direct higher patent classification code of the main patent classification code of 1 dot level in C5 level at a left side, and inputs the patent classification code of the main group level in C4 level, and those procedures are performed up to the topmost level in a series. The same manner performed for the main IPC is applied to at least one sub- IPC included in the patent document. At this time, other bibliographical data, such as the applicant, can be obtained through the document number, so that they are the selective elements.

Korean Patent Application No. 10-2005-0111868 is assigned H04B 7/26 and H04B 7/15 dated of January, 2006. Based on this, it will be exemplari Iy described through Table 60 (for convenience of the marking, the sub-class part of the patent classification code below the main group will be omitted.) [Table 60]

If there is two or more sub IPCs, they are identically processed parallel Iy in a right side of the information relating to H04B 7/15, as shown in Table 61 below.

Korean Patent Application No. 10-2006-0012606 includes the patent classification code of H04B 7/04, H04B 7/155, H04Q 7/30 assigned in January, 2006, based on which, it will be exemplari Iy described (for convenience of the marking, AppName, etc. is omitted and H04Q 7/30- relation is represented in a lower line.) [Table 61]

In Table 62 below, Korean Patent Application No. 10-2005-0042032 includes the patent classification codes of H04B 7/02 and H04B 7/14. [Table 62]

It is obvious to one skilled in the art that the multi-dimensional analysis operation result table data can be generated with respect to USPC, FT, etc. by the multi-dimensional analysis operation result table generating module 402. The multi-dimensional analysis operation result table generating module 402 generates multi-patent classification code pair information as included in Table 63 below from the series of information of the level- based multi-patent classification codes. The types of multi-patent classification code pair has been described in the homogeneous patent classification code correlation pre-processing module. In table 63 below, the bibliographical particular including only one information, such as the application date and registration date, can be created and included in the side. However, as described above, a separate row must be created in the case of join applicants or join inventors. That is, an identical record in which only the applicant or inventor is different is generated, and based on this information, it is possible to perform the applicant-based convergence analysis and inventor-based convergence analysis. In order to discuss the present model, let us assume that there are only three patent documents including Korean Patent Application No. 10-2005-0111868, Korean Patent Application No. 10-2006-0012606, and Korean patent application No. 10-2005-0042032 (of course, by repeatedly processing all the obtained documents, it is possible to generate the following data in relation to all the patent classification code combinations (Ai, Bj) based on all the documents). Here, for the patent classification code combination (Ai, Bj), it is expected to be possible to create the table as follows. Korean Patent Application No. 10-2005-0111868 includes 4 pairs of patent classification code combinations, Korean Patent Application No. 10-2006-0012606 includes 22 pairs (6+16) of patent classification code combinations (main IPC pair, sub-IPC pair) and 20 pairs of patent classification code combinations (sub-IPC pair, sub-IPC pair), and Korean patent application No. 10-2005-0042032 includes 1 pair of patent classification code combinations, so that three patent applications includes 47 pairs of patent classification code combinations. It is of course possible to represent all the pairs, but for the convenience of the description, the present inventive idea will be described with Table 63 in which a certain part of the pairs is omitted, and this can be corresponding to total 47 pairs. [Table 63]

When the data is generated in the scheme of the patent classification code combination (Ai, Bj), there may occur an error in sorting, index processing, and/or roll-up operation. Therefore, it is more preferable to conceptually process the patent classification code combination as shown in Table 64 below, and the multi-dimensional analysis operation result table generating module 402 can generate multiple patent classification code pair information as follow. [Table 64]

According to the object of the statistic processing and analysis, it is more preferable that the multi-dimensional analysis operation result table generating module 402 generates the multiple patent classification code pair information as shown in Table 65 below (for the convenience of description, the individual document number is not indicated, but is completely identical to the sequence of the document number shown in Table 64 above. Further, it is preferred that the multi-dimensional analysis operation result table data as shown in Table 65 below generated by the multi-dimensional analysis operation result table generating module 402 co-includes other bibliographical particulars (the applicant, inventor, application date, registration date) of the document number (Below, M/S is a field implying that if the patent classification code is obtained from the main IPC, indicated M, and if the patent classification code is obtained from the sub IPC, indicate S. This is because the reference of the convergence is identified on a main IPC, preferably.) [Table 65]

If there is the multi-dimensional analysis operation result table data including the above convergence information for all the patent documents including two or more patent classification codes, it is possible to extract/calculate the following information.

First one is a subject information non-utilization convergence analysis when the patent classification code is obtained. When a single patent classification code is obtained or given, it is possible to search the ranking of the patent classification code well converging (having high frequency of being made as one of the pair) with the patent classification code. For example, in the above table, it is possible to obtain the IPCs having the high convergence frequency in the respective IPC levels, such as the patent classification code well converging with H04B 7/04 is H04B 7/155C7 in CP(IPC 3 dot level) level and H04B 7/15 in C6 level, and if the statistic/analysis information is generated with respect to the obtained IPCs (through processing by the group-by, count, and rank command), it is possible to identify the ranking information on the IPCs having the high convergence frequency (for example, it is possible to generate the ranking information relating to the convergence, such as "the most converging IPC of C6 level with H04B 7/04 is H04B 7/15 and IPC of C6 level converging with H04B 7/04 by the next ranking is H04Q 7/30". Of course, the specific document in which the convergence is created can be specifically identified through the document number, but by querying the IPC pair creating the convergence to the search engine or DBMS 291, it is possible to the corresponding documents. It is possible to obtain year-based/term- based statistic/analysis information on the patent classification codes well converging with the obtained patent classification code according to the respective IPC levels (IPC sub-class, main group, 1 dot, 2 dot, ... ), which can be implemented by searching the obtained patent classification code (e.g. H04B 7/04) in one side of the table (preferably in main IPC), determining the level in every record (row) Cl ~ Cn including the searched patent classification code (in which the column is determined), and generating the statistic/analysis data (through the group by, count, and rank command) with respect to every patent classification code existing in the determined level. Further, if the term/date information is included in the multi-dimensional analysis operation result table data, it is possible to generate the statistic/analysis data of the patent classification codes well-converging with the obtained patent classification code according to the term (e.g. according to the year).

Claims

[CLAIMS]

[Claim 1]

A method of providing patent information by a system providing patent information to a computer of a user, the method comprising the steps of: obtaining selection information for a patent document set; specifying a to-be-processed patent document set based on the selection information! and generating processed patent information by performing at least one preset processing for the specified to-be-processed patent document set, wherein at least one patent classification code is used in at least one of the step of specifying the to-be-processed patent document set and the step of generating the processed patent information by performing said at least one preset processing, use of the patent classification code corresponds to processing of information comprising patent information of a first patent classification code and lower patent classification codes of the first patent classification code when the first patent classification code is given in at least one of the step of obtaining the selection information and the step of generating the processed patent information and corresponds to a preset level in a patent classification code system to which the first patent classification code belongs, and the patent classification code belongs to at least one of International Patent Classification (IPC) system, United States Patent Classification (USPC) system, File Index (FI) system of Japan, F-term (FT) system of Japan, and European Classification (ECLA) system.

[Claim 2]

The method of claim 1, wherein the preset level is a dot level patent classification code, and the dot level patent classification code comprises a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code in the patent classification code system.

[Claim 3]

The method of claim 2, wherein the dot level patent classification code comprises a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subgroup or below when the dot level patent classification code belongs to the IPC system, comprises a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subclass or below when the dot level patent classification code belongs to the USPC system, comprises a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subgroup or below when the dot level patent classification code belongs to the FI system of Japan, comprises a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes under a theme code when the dot level patent classification code belongs to the F-term(FT) system of Japan, and comprises a dot or dots in title information thereof, which is a description corresponding to the dot level patent classification code from among patent classification codes of a subgroup or below when the dot level patent classification code belongs to the ECLA system.

[Claim 4]

The method of claim 1, wherein, in the processing of information comprising patent information of the first patent classification code and lower patent classification codes of the first patent classification code, when the first patent classification code is a dot level patent classification code, at least one patent document in a patent document set including all lower patent classification codes of the dot level first patent classification code is subjected to information processing.

[Claim 5]

The method of claim 1, wherein the selection information is obtained by at least one of: 1) an obtainment method of receiving input of the selection information through a search formula from the computer of the user; 2) an obtainment method of receiving selection information, which is selected from bibliographical information included in the preset patent document presented in the computer of the user and is transmitted from the computer of the user; 3) an obtainment method of receiving selection information on a particular patent document set selected in at least one patent document set generated in advance by the user; and 4) an obtainment method of receiving at least one preset condition from the system.

[Claim 6]

The method of claim 5, wherein, in the step of specifying the to- be-processed patent document set, when one of the obtainment methods 1) to 4) has been executed, at least one first patent classification code is included in the selection information during execution of the methods 1) to 4), and the first patent classification code is a dot level patent classification code, a patent document set including all lower patent classification codes of the dot level first patent classification code is specified as the to-be-processed patent document set.

[Claim 7] The method of claim 6, wherein a process of specifying the patent document set including all lower patent classification codes of the dot level first patent classification code as the to-be-processed patent document set comprises the steps of: obtaining all lower patent classification codes of the dot level first patent classification code; and specifying the to-be-processed patent document set by using a search formula or query formula including the first patent classification code and all the lower patent classification codes thereof.

[Claim 8] The method of claim 1, wherein individual patents of the patent document set include at least one patent granted a weight according to at least one preset weight granting policy, and the processed patent information is generated while reflecting the weight granted to the individual patent.

[Claim 9]

The method of claim 8, wherein the weight granting policy comprises calculation of at least one of 1) preset counting information and 2) preset calculation value information in relation to the individual patent , a counting value for the preset counting information comprises at least one of 1) a number of families in at least one of a first nation and a second nation, 2) a number of claims, and 3) a cited-by number, and the preset calculation value information comprises at least one of

1) at least one evaluation score for an inventor of the patent, 2) at least one evaluation score for an applicant of the patent, 3) at least one preset calculation value for a patent classification code included in the patent of the inventor, 4) at least one preset calculation value for a patent classification code included in the patent of the applicant, 5) at least one preset calculation value reflecting financial information of the applicant, and 6) at least one preset calculation value for the patent classification code itself included in the patent.

[Claim 10]

The method of claim 1, wherein processed patent information corresponding to the patent classification code information is provided together when the processed patent information is provided to the computer of the user, and the processed patent information provides processed patent information corresponding to at least one direct lower patent classification code of the patent classification code in a patent classification code system including the patent classification codewhen there is a request for the processed patent information corresponding to at least one direct lower patent classification code of the patent classification code from the computer of the user.

[Claim 11] A method of providing patent information by a system providing patent information to a computer of a user, the method comprising the steps of: obtaining selection information for a patent document set; specifying a to-be-processed patent document set based on the selection information; and generating processed patent information by performing at least one preset processing for the specified to-be-processed patent document set, wherein the selection information is obtained by at least one of:

1) an obtainment method of receiving input of the selection information through a search formula from the computer of the user; 2) an obtainment method of receiving selection information, which is selected from bibliographical information included in the preset patent document presented in the computer of the user and is transmitted from the computer of the user! 3) an obtainment method of receiving selection information on a particular patent document set selected from at least one patent document set generated in advance by the user; and 4) an obtainment method of receiving at least one preset condition from the system, the processed patent information comprises at least one piece of multi-dimensional analysis information generated through at least one preset analysis of the specified to-be-processed patent document set, dimensions of the multi-dimensional analysis information comprise at least one of 1) time, 2) patent classification code, 3) applicant, 4) inventor, 5) agent, 6) nation, 7) keyword, and 8) convergence pattern, the multi-dimensional analysis information comprises at least one preset counting value for at least one of the dimensions, a calculation value obtained by at least one preset calculation formula, a query result value obtained by at least one query formula, and a result value of processing by at least one analysis processing algorithm, for the to-be- processed patent document set .

[Claim 12]

The method of claim 11, wherein, for patent classification codes included in the to-be-processed patent document set, the analysis information is generated by applying at least one of 1) an analysis method for analysis of all patent classification codes and 2) an analysis method for analysis of only main patent classification codes, and the analysis method for analysis of all patent classification codes generates analysis information by applying at least one of a) an analysis method of granting a uniform weight to all patent classification codes, and b) an analysis method of granting different weights to main patent classification codes and sub-patent classification codes.

[Claim 13]

The method of claim 11, wherein the analysis is an analysis of convergence between patent classification codes, which comprises at least one of association analysis, frequent occurrence pattern analysis, correlation analysis, and grouping between at least two patent classification codes included in patent documents including at least two patent classification codes of an identical patent classification system from among the patent documents included in the to-be-processed patent document set.

[Claim 14]

The method of claim 13, wherein, in the analysis of convergence between patent classification codes, when said at least two patent classification codes of the identical patent classification system are included in a single patent document and when said at least two patent classification codes are not highest patent classification codes or patent classification codes of a preset level, the analysis of convergence between patent classification codes is performed for each of said at least two patent classification codes together with higher patent classification codes thereof up to a preset level of each of said at least two patent classification codes.

[Claim 15]

The method of claim 11, wherein the analysis of convergence comprises at least one of 1) a convergence analysis between at least two technical keywords generated based on text information extracted from at least one preset field of one patent document, 2) a convergence analysis between co-inventors of a patent document having co-inventors, 3) a convergence analysis between co-applicants of a patent document having co- applicants, and 4) a convergence analysis between citation patent documents included in citation information of a single patent document, for individual patent documents of the to-be-processed patent document set, and the convergence analysis comprises at least one of association analysis, frequent occurrence pattern analysis, correlation analysis, and grouping between at least two keywords, co-inventors, co-applicants, and co-citation patent documents included in the patent documents.

[Claim 16]

The method of one of claims 13 to 15, further comprising the step of visualizing a result of the convergence analysis through a network diagram including at least two nodes, wherein the visualized result comprise at least one of information on the nodes and information on lines interconnecting the nodes.

[Claim 17]

The method of claim 16, wherein the information on the lines interconnecting the nodes comprises numerical information, and the method further comprises the steps of: obtaining selection information on the numerical information or the lines from the computer of the user; extracting root patent document information corresponding to the numerical information from a patent information database including the to- be-processed patent document set; and transmitting the extracted root patent document information to the computer of the user.

[Claim 18]

The method of claim 1, further comprising the steps of: obtaining change information of the to-be-processed patent document set from the computer of the user; and specifying the to-be-processed patent document set based on the obtained change information, wherein the obtaining of the change information comprises: providing an interface for the obtaining of the change information to the computer of the user! and obtaining selection information of the user for the interface, the interface comprises at least one of 1) time type selection interface, 2) time interval selection interface, 3) nation selection interface 4) patent classification code selection interface 5) citation specification interface, 6) family information-including selection interface, and 7) after-search selection interface, the time type selection interface comprises a selection interface providing selection of at least one of application date, publication date, and registration date, the time interval selection interface comprises a selection interface providing selection of at least one of year/month/week/day of each of a starting period and an ending period, the nation selection interface comprises a selection interface providing selection of at least one between at least two nations, the patent classification code selection interface comprises a selection interface providing selection of at least one between a patent classification code type selection interface, a patent classification code level selection interface, and a main patent classification code and sub- patent classification code selection interface, the citation specification interface comprises a selection interface providing selection of at least one of a citation direction interface for selection between a backward citation and a forward citation, a citation depth interface for selection of a citation depth, a citation selection interface for selection between a direct citation and an indirect citation, the family information-including selection interface comprises a selection interface for at least one of determination if family information is included and selection of a nation range, and the after-search selection interface comprises a processing interface for presenting a preset field for search in order to find the to-be-processed patent document set, receiving a search phrase through the preset field from the computer of the user, providing a search result corresponding to the search phrase, receiving at least one selection information selected in the search result, and determining only a patent document set relating to the selection information as the to-be-processed patent document set.

[Claim 19]

The method of claim 11, wherein the analysis is a set-based citation analysis that treats at least two patent documents as one analysis unit, the set-based citation analysis comprises at least one of 1) backward citation patent document set analysis, 2) forward citation patent document set analysis, and 3) citation occurrence self patent document set analysis, the analysis of .1) to 3) creates at least one preset multi^¬ dimensional analysis information from the patent document set, which is an object of the analysis of 1) to 3), a backward citation patent document set specified for the backward citation patent document set analysis comprises a document set including citation information of at least one citation depth included in the to-be- processed patent document set , a forward citation patent document set specified for the forward citation patent document set analysis comprises a later-filed document set including citation information of at least one citation depth citing the to-be-processed patent document set , a citation occurrence self patent document set specified for the citation occurrence self patent document set analysis comprises a patent document set comprising patent documents cited at least one time from among the patent documents of the to-be-processed patent document set.

[Claim 20]

The method of claim 19, wherein, in the set-based citation analysis, specification of the backward citation patent document set comprises specification of the patent document set and an indirect backward citation document set of the patent document set and specification of the forward citation patent document set comprises specification of the patent document set and an indirect forward citation document set of the patent document set , the indirect backward citation document set comprises patent documents, which have application dates prior to application dates of to- be-processed patent documents of the to-be-processed patent document set and commonly cite patent documents cited by the to-be-processed patent documents, and the indirect forward citation document set comprises patent documents, which have application dates later than application dates of to-be-processed patent documents of the to-be-processed patent document set and commonly cite patent documents cited by the to-be-processed patent documents.

[Claim 21] The method of claim 11, wherein the multi-dimensional analysis information is transmitted in a form of a table including numerical information to the computer of the user, and the method further comprises the steps of: receiving selection information on the numerical information from the computer of the user; extracting root patent document information corresponding to the numerical information from a patent information database comprising the to-be-processed patent document set ; and transmitting the extracted root patent document information to the computer of the user.

[Claim 22]

The method of claim 11, wherein at least one of the step of specifying the to-be-processed patent document set and the step of generating the processed patent information by performing at least one preset processing uses at least one patent classification code, use of the patent classification code comprises information processing of a first patent classification code together with lower patent classification codes of the first patent classification code when the first patent classification code is given in at least one of the step of receiving the selection information and the step of generating the processed patent information and corresponds to a preset level in a patent classification code system to which the first patent classification code belongs, and the patent classification code system comprises at least one of the IPC, USPC, FI and FT of Japan, and ECLA.

[Claim 23]

The method of claim 11, wherein the obtainment method of receiving selection information, which is selected from bibliographical information included in the preset patent document presented in the computer of the user and is transmitted from the computer of the user, comprises the steps of: providing at least one of applicant information including at least one applicant, patent classification code information including at least one patent classification code, inventor information including at least one inventor, and agent information including at least one agent to the computer of the user; obtaining selection information on at least one of the applicant information, the patent classification code information, the inventor information, and the agent information from the computer of the user; and generating a preset multi-dimensional analysis result by processing a patent document set specified by the selection information as a to-be-processed patent document set, and providing the generated preset multi-dimensional analysis result to the computer of the user.

[Claim 24] The method of claim 23, wherein the system generates at least one medium multi-dimensional analysis result table for easily and rapidly creating a multi-dimensional analysis result according to the applicants, the patent classification codes, the inventors, or the agents, and the method further comprises the steps of: providing at least one of at least one preset applicant-based mult i-dimensional analysis list or menu, at least one preset patent classification code-based multi-dimensional analysis list or menu, at least one preset inventor-based mult i-dimensional analysis list or menu, and at least one preset agent-based multi-dimensional analysis list or menu to the computer of the user; obtaining selection information on the multi-dimensional analysis list or menu from the computer of the user; and when an analysis list of the obtained selection information is obtainable from the medium multi-dimensional analysis result table, generating an analysis result of the obtained selection information by referring to the medium multi-dimensional analysis result table and providing the generated analysis result to the computer of the user.

[Claim 25]

A method of providing patent information by a system providing patent information to a computer of a user, the method comprising the steps of: obtaining patent document information on at least one patent document published in a first nation; when more than one year has passed after an application date of the published patent document in the first nation and a date on which the patent document information is obtained, referring to at least one of 1)

PCT application information of the published patent document, 2) second nation family information claiming a priority to a second nation based on the published patent document of the first nation, and 3) current status information of a second nation patent document corresponding to the second nation family information; determining if the published patent document corresponds to a first case in which the published patent document has no PCT application information and no second nation family information, a second case in which the published patent document has PCT application information but has no second nation family information, and the application date of the published patent document in the first nation corresponds to a date after passage of a preset term from the date on which the patent document information is obtained, or a third case in which the published patent document has PCT application information and second nation family information, and the current status information of the second nation patent document corresponds to at least one of cancellation, expiration, final rejection, and final invalidation! and when the published patent document corresponds to one of the first case, the second case, and the third case, transmitting patent document information of the first nation and preset display information to the computer of the user, wherein the preset display information indicates that the published patent document corresponds to one of the first case, the second case, and the third case and the preset display information is displayed around the patent document information of the first nation when the patent document information of the first nation is displayed, wherein the published patent document has at least one second nation.

[Claim 26] A patent information system comprising: a patent document master DB; a patent classification code master DB; and a lower patent classification code pre-processing module, wherein the lower patent classification code pre-processing module performs processing of information comprising patent information of a first patent classification code and lower patent classification codes of the first patent classification code when the first patent classification code is given and corresponds to a preset level in a patent classification code system to which the first patent classification code belongs.

[Claim 27] The patent information system of claim 26, further comprising at least one of a modified search formula generation module and a modified query formula generation module, the modified search formula generation module and the modified query formula generation module generate a search formula or a query formula including the first patent classification code and all the lower patent classification codes of the first patent classification code when they have obtained all the lower patent classification codes of the first patent classification code including dot level lower patent classification codes thereof.

[Claim 28]

The patent information system of claim 26, further comprising a weight granting module, wherein the weight granting module grants a weight in at least one point of view to each of individual patents of a specified patent document set according to at least one preset weight granting policy.

[Claim 29]

The patent information system of claim 28, wherein the weight granting policy comprises calculation of at least one of 1) preset counting information and 2) preset calculation value information in relation to the individual patent, a counting value for the preset counting information comprises at least one of 1) a number of families in at least one of a first nation and a second nation, 2) a number of claims, and 3) a cited-by number, and the preset calculation value information comprises at least one of 1) at least one evaluation score for an inventor of the patent, 2) at least one evaluation score for an applicant of the patent, 3) at least one preset calculation value for a patent classification code included in the patent of the inventor, 4) at least one preset calculation value for a patent classification code included in the patent of the applicant, 5) at least one preset calculation value reflecting financial information of the applicant, and 6) at least one preset calculation value for the patent classification code itself included in the patent.

[Claim 30] The patent information system of claim 26, further comprising a patent analysis intelligence module for at least one patent analysis of a specified to-be-analyzed patent document set, wherein the patent analysis intelligence module generates analysis information from patent classification codes included in the to-be- processed patent document set by applying at least one of 1) an analysis method for analysis of all patent classification codes and 2) an analysis method for analysis of only main patent classification codes, the analysis method for analysis of all patent classification codes generates analysis information by applying at least one of a) an analysis method of granting a uniform weight to all patent classification codes, and b) an analysis method of granting different weights to main patent classification codes and sub-patent classification codes, the to-be-analyzed patent document set is specified by using at least one of: 1) an obtainment method of receiving input of the selection information through a search formula from the computer of the user; 2) an obtainment method of receiving selection information, which is selected from bibliographical information included in the preset patent document presented in the computer of the user and is transmitted from the computer of the user; 3) an obtainment method of receiving selection informationon a particular patent document set selected from at least one patent document set generated in advance by the user! and 4) an obtainment method of receiving at least one preset condition from the system, the processed patent information comprises at least one piece of multi-dimensional analysis information generated through at least one preset analysis of the specified to-be-processed patent document set, dimensions of the multi-dimensional analysis information comprise at least one of 1) time, 2) patent classification code, 3) applicant, 4) inventor, 5) agent, 6) nation, 7) keyword, and 8) convergence pattern, the multi-dimensional analysis information comprises at least one preset counting value for at least one of the dimensions, a calculation value obtained by at least one preset calculation formula, a query result value obtained by at least one query formula, and a result value of processing by at least one analysis processing algorithm, for the to-be- processed patent document set . [Claim 31] The patent information system of claim 30, wherein the analysis is an analysis of convergence between patent classification codes, which comprises at least one of association analysis, frequent occurrence pattern analysis, correlation analysis, and grouping between at least two patent classification codes included in patent documents including at least two patent classification codes of an identical patent classification system from among the patent documents included in the to- be-processed patent document set . [Claim 32] The patent information system of claim 31, wherein, in the analysis of convergence between patent classification codes, when said at least two patent classification codes of the identical patent classification system are included in a single patent document and when said at least two patent classification codes are not highest patent classification codes or patent classification codes of a preset level, the analysis of convergence between patent classification codes is performed for each of said at least two patent classification codes together with higher patent classification codes thereof up to a preset level of each of said at least two patent classification codes. [Claim 33]

The patent information system of claim 31, wherein the analysis of convergence comprises at least one of 1) a convergence analysis between at least two technical keywords generated based on text information extracted from at least one preset field of one patent document, 2) a convergence analysis between co-inventors of a patent document having co-inventors, 3) a convergence analysis between co-applicants of a patent document having co-applicants, and 4) a convergence analysis between citation patent documents included in citation information of a single patent document, for individual patent documents of the to-be-processed patent document set, and the convergence analysis comprises at least one of association analysis, frequent occurrence pattern analysis, correlation analysis, and grouping between at least two keywords, co-inventors, co-applicants, and co-citation patent documents included in the patent documents. [Claim 34] A patent information system comprising: a patent document master DB; a patent classification code master DB; and a patent analysis intelligence module for at least one patent analysis of a specified to-be-analyzed patent document set, the analysis is a set-based citation analysis that treats at least two patent documents as one analysis unit, the set-based citation analysis comprises at least one of 1) backward citation patent document set analysis, 2) forward citation patent document set analysis, and 3) citation occurrence self patent document set analysis, the analysis of 1) to 3) creates at least one preset multi^¬ dimensional analysis information from the patent document set, which is an object of the analysis of 1) to 3), a backward citation patent document set specified for the backward citation patent document set analysis comprises a document set including citation information of at least one citation depth included in the to-be- processed patent document set , a forward citation patent document set specified for the forward citation patent document set analysis comprises a later-filed document set including citation information of at least one citation depth citing the to-be-processed patent document set , a citation occurrence self patent document set specified for the citation occurrence self patent document set analysis comprises a patent document set comprising patent documents cited at least one time from among the patent documents of the to-be-processed patent document set . [Claim 35]

The patent information system of claim 34, wherein, in the set- based citation analysis, specification of the backward citation patent document set comprises specification of the patent document set and an indirect backward citation document set of the patent document set and specification of the forward citation patent document set comprises specification of the patent document set and an indirect forward citation document set of the patent document set, the indirect backward citation document set comprises patent documents, which have application dates prior to application dates of to- be-processed patent documents of the to-be-processed patent document set and commonly cite patent documents cited by the to-be-processed patent documents, and the indirect forward citation document set comprises patent documents, which have application dates later than application dates of to-be-processed patent documents of the to-be-processed patent document set and commonly cite patent documents cited by the to-be-processed patent documents. [Claim 36]

A patent information system comprising: a patent document master DB; a pre-processing module; and a patent information processing basis module, wherein the patent information processing basis module comprises a multi-dimensional analysis calculation result table generation module and at least one table generated by the multi-dimensional analysis calculation result table generation module, the patent information system generates at least one medium multi- dimensional analysis result table for easily and rapidly creating a multi^¬ dimensional analysis result according to the applicants, the patent classification codes, the inventors, or the agents, and the patent information processing basis module performs: providing at least one of at least one preset applicant-based multi-dimensional analysis list or menu, at least one preset patent classification code-based multi-dimensional analysis list or menu, at least one preset inventor-based multi-dimensional analysis list or menu, and at least one preset agent-based multi-dimensional analysis list or menu to the computer of the user! obtaining selection information on the multi-dimensional analysis list or menu from the computer of the user; and when an analysis list of the obtained selection information is obtainable from the medium multi-dimensional analysis result table, generating an analysis result of the obtained selection information by referring to the medium multi-dimensional analysis result table and providing the generated analysis result to the computer of the user.