JP2002245063A - Method for managing and retrieving database - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for managing and retrieving a database for easy classification and retrieval in a data managing system which is constructed by directing through the use of a data characteristic without necessity to previously setting a preservation structure. SOLUTION: The method consists of a management part and a retrieval part. In the management part, an attribute is defined concerning data to be preserved, a relation attribute related to the data is searched from the definition so as to form subordinate relation, data is preserved at a specific position by a specific form based on the attribute and the relation attribute and each attribute number in connection with the position of the file is displayed and related to the defined attribute. In the retrieval part, the attribute of a condition which is designated by an attribute recording segment is detected, document data concerning the attribute which is designated by a file recording segment is retrieved and document data of the attribute which is designated by the file recording segment with respect to the relation attribute is retrieved.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database management and search method, and more particularly to a method in which a user actively defines data into a plurality of different attributes. The present invention relates to a database management and a search method that can be easily stored and that can quickly and accurately retrieve data because the definition is familiar to the user.

[0002]

2. Description of the Related Art With the formation of an information-oriented society, the collection and application of information from individuals to corporations and organizations are extremely important issues. In particular, for a newspaper company or a publisher, the accumulation and storage of data is an important commercial resource. Therefore, this information must be managed in a specific way. The initial data management method merely stores the actual information and stores useful information in a concentrated manner. At the time of use, the data is narrowed down by turn. Such traditional methods of managing data do not form any structure and are not adaptable to modern society where efficiency and speed are required.

[0003] In recent years, computers have become widespread and widely used for data management. Computer data management is more efficient than traditional data management methods. The data management method by the computer is generally classified into a method of classifying data by an index and a system of classifying data by using a multiple index similar to the use of a card index in a reference list. In the former method, a data manager subjectively judges data based on a predetermined classification standard. By such subjective judgment, each data is stored in a file to which the data belongs, and when necessary, a file having a specific property is searched one by one to obtain necessary data. In the latter, the document data is defined and stored based on fixed columns such as an author, a description date, a described publication, a headline, or a classification number. When searching the stored document data, the search proceeds according to the defined property of the document data.

However, both the method of classifying by the index and the method of classifying by the multiple index similar to the card index have disadvantages. The index classification method belongs to the human classification method, and is entirely based on the administrator's subjective.
Therefore, when there is an error in the recognition of the property of the data by the manager and the predetermined standard, a situation occurs in which the data of the type A is stored as the data of the type B. Since this kind of classification method roughly classifies data into a single class, if data is recursively classified into the wrong class, the data to be stored in the original classification will be lost. Therefore, relying solely on classification by a single category is clearly extremely dangerous.

[0005] Further, in the method of classifying by the multiple index similar to the card index, since a plurality of fixed columns can be used for retrieval, the classification is unstable compared to the classification method by the single classification. Factors can be reduced. However, it has inherent defects. The obvious drawback is the fixed column. That is, a fixed column cannot meet all the needs of a scrapbook. For example, in a database of a newspaper company or a publisher, since the data categories are wide, it is not possible to appropriately define different data of different categories using only a few columns. As an example, assume that there are seven authors of certain document data. However, only three authors can be stored in the author column in the database management system. in this case,
You must select the author to be entered in the author field. That is, only three authors can be registered in the document data, and the other four must be discarded.
Therefore, when searching by the author's name, a blind spot in data search is formed. In addition, the number of authors that can be registered is not only limited, but a similar problem may occur with respect to headings, time, or related names, cases, times, place names, articles, and the like.

In addition to the data management method described above, there is a method of searching for data using a keyword. This keyword is used to index data, and by inputting a keyword related to a sentence to be searched, the data management system searches the entire sentence one by one based on the input keyword, and searches the related sentence. Find out.
However, such a search method is clearly lacking in accuracy. That is, since the keyword is used as a basis for the determination in the search, all documents including the keyword are searched as a result of the detection, and sentences that are not related to the meaning of the keyword are also searched.
Such a method is inefficient in a data search operation.

[0007]

SUMMARY OF THE INVENTION The present invention relates to a data management system which does not require a storage structure to be set in advance, and is constructed by using the data itself as a starting point and orienting it according to the characteristics of the data. It is an object of the present invention to provide a database management and search method that allows efficient search.

Another object of the present invention is to provide a database management and search method that is easy for an administrator to understand and that can be easily classified and searched.

[0009]

According to the present invention, a system is constructed by determining the direction according to the characteristics of the data itself. That is, without providing in advance a plurality of fields for defining the characteristics of the document data,
A column is created based on all information that is actively referred to by the administrator, and a plurality of definitions are made for document data using the column that has been actively added. In other words, cross-concentration or connection-concentration calculations are performed to facilitate database search.

The database management and search method according to the present invention includes an attribute recording segment and a file recording segment. The attribute recording segment is
It is for recording the attributes that the administrator defines for different document data, and each attribute is divided into items such as "Attribute name", "Attribute number", and "Related attribute number" Record. The "attribute name" is actively input by the administrator, and the "attribute number"
Is a number automatically generated each time an attribute is recorded by the recording management unit of the attribute recording segment, and the “attribute number” may be a continuous number. The “related attribute number” is input by the administrator and is another attribute number related to the attribute of the document data to be stored, and the “related attribute number” is a part of the attribute of the document data. It is said that.

In the file recording segment, document data defining an attribute is recorded and stored. A file number, an address to save, and an attribute number are recorded for each document data to be saved. The file number is used for identification in the same manner as the attribute number, and it is sufficient that the file number does not overlap with the attribute number. The database is structured by such a database management and search method to achieve a user-friendly environment.

When the database management and search method is executed by a computer, a disk unit name, a root, and a file name are displayed at the address of the file recording segment.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION In order to explain the object and technology of the present invention, specific examples will be given and described in detail with reference to the drawings. The database management and search method according to the present invention can be divided into a management part and a search part. For the management part, starting from a user's point of view, defining an attribute for the data to be stored, and searching for a related attribute related to the data from the defined attribute to form a dependency. Storing the data in a specific format based on the attribute and related attributes, storing the data in a specific location, and displaying the attribute number associated with the location of the file, respectively, Associating with an attribute.

The step of defining the attribute is a step of setting the direction from the user's own viewpoint and defining the attribute for the document data to be stored. In other words, when the user defines the attributes of the document data, the items of the attributes that define the information by recognizing the information that the user subjectively deserves reference without being restricted by fixed fields or normal forms. Listed in
For example, one letter, one sign, a specific noun, a number, a person's name or an organization. For example, to save a sentence on the 1990 Taiwan-Canada trade situation,
Any of "1990", "Taiwan", "Canada", "Trade", "Excess import", etc. can be defined as attribute items. Such attributes are actively submitted by the user and match the user's point of view, so when the user performs a search, the attributes can be accurately pointed out and the data required quickly Can be obtained. Also, the attribute items are actively submitted by the user, and there are no restrictions on the items in the columns. Therefore, in the conventional card index system, there is no disadvantage that the document file cannot be defined in detail due to the limitation of the column items.

The step of defining the related attributes is a step of structuring the data when the data to be stored and managed is accumulated and the number of items of the defined attributes increases accordingly. This is to newly define a related attribute related to the attribute from the existing attribute items, thereby forming a dependent relationship between the attributes and structuring the data.

FIG. 1 discloses a data format of the attribute item. As shown in the figure, the style includes an attribute number, an attribute name, and a related attribute number, and the defined names include “Earth”, “Asia”, “America”, “Taiwan”, “Taichung (Taiwan ) "," History "," Canada ", and the like, and numbers from 0001 to 0007 are assigned in order. In addition, the names of the attributes form a specific dependency with the names of the other attributes. That is, since the attribute "Asia" is a part of the earth, the related attribute number of "Asia" is 0001. Also, since the attribute "America" is also a part of the earth, the related attribute number of "America" is also 00.
01. Since the attribute “Taiwan” is a part of Asia, the related attribute number of “Taiwan” is 0002. Since the attribute “Taichung” is a part of Taiwan, the related attribute number of “Taichung” is 0004. Since the attribute “Canada” is a part of the Americas, the related attribute number of “Canada” is 0003. However, the related attribute numbers are not recorded in the attribute items of “Earth” and “History”. This indicates that the attribute is not included in the dependent range of the other attribute. Assigning these attribute-related numbers is intended mainly to achieve structuring of data based on the relevance between data and facilitate retrieval.

The step of saving the attribute is a step of saving the attribute item whose definition has been completed in a specific format. The attribute recording segment is mainly constructed and performed. A name, an attribute number, and an attribute-related number (see FIG. 1) are recorded in the attribute recording segment. The name of the attribute is input actively by the user when the user defines the attribute. is there. The attribute number is automatically generated by the attribute record management unit in the attribute record segment. In principle, each attribute number must be unique and must not overlap with other attribute item numbers. For this reason, the easiest method is to display the numbers consecutively in the order before and after saving. The attribute-related number is for searching for and recording an attribute item related to the stored data based on the attribute item already defined by the user.

The step of storing the data is a step of constructing and executing a file recording segment. In the file recording segment, a file number, a storage position, an attribute number of document data to be stored, and the like are recorded. The file number is a number assigned to the data to be stored, is used for identification like the attribute number, and is a unique number and must not be a duplicate number.

The storage location mainly indicates the location where the file is to be stored. If the method of the present invention is executed by a computer, the storage location includes the name of the disk unit, the root, and the location of the file. Displays the name. For example, as disclosed in FIG. 2, the name of the disk unit, the root, and the name of the file are recorded at the position where each document data is stored.

The attribute number is actively defined by the user, and is recorded in the file recording segment along with the constructed file itself in addition to being recorded in the attribute recording segment.

FIG. 2 discloses the contents of the file recording segment. As shown, C: \ Document \ Artricle. Do
The data stored in location c is Taichung, history,
It is defined by three attribute items such as Canada. Therefore, the document data is 006, 005,
007 including the attribute number. The attribute of the data may be changed due to the modification of the content.
In this case, the contents can be corrected by increasing or decreasing the attribute number in the file recording segment. For example, in the content stored in FIG. 2, the storage location is D: \ Person \ Report. It has document data that is Xls. The attribute number associated with the data is only Taiwan (0004). If the content of the data is modified, the definition of other attributes must be increased. In this case, it can be dealt with by adding a record of a related attribute to the data (see FIG. 3).
The data constitutes the relationship between history and attributes, in addition to the relationship between Taiwan and attributes. Conversely, if the record disclosed in FIG. 3 is deleted, the data will no longer be associated with a historical attribute item.

The above has clearly stated the principle and purpose of the data classification of the present invention. The search method consists of a “specified process” in which the user specifies attributes, a “related attribute search process” in which attributes of specified conditions are detected from the attribute recording segment, and an attribute related to the attribute specified from the file recording segment. "Searching for data related to the specified attribute" to search for the document data to be performed, and "Searching for the document data related to the attribute related to the specified attribute from the file recording segment" Searching for related data ".

The "designating" step is performed by the user, and includes the following two types of methods. That is, one is a method of browsing the material of the attribute recording segment and designating the attribute. Each record includes an attribute number and a related attribute number. Taking FIG. 1 as an example, seven attribute items are described in the figure,
If the attribute items without the related attribute item of the subordinate relation are listed first, both the earth item and the history item are included. An analysis of the attributes of the earth shows that the Americas and Asia are related attributes. Also, if you choose Asia and the Americas,
If you choose Asia, Taiwan is a relevant attribute of Asia. Therefore, Taiwan can be directly used as the designated attribute for searching. By such a method of selecting items one by one, related attributes can be quickly found.

Another method is to perform the collation one by one. That is, this is a method in which the specified attribute item and the attribute item recorded in the attribute recording segment are collated one by one. Speaking of the processing speed of the current computer, it is possible to perform an extremely quick and easy matching of the attribute item with all the attribute items recorded in the attribute recording segment one by one.
Therefore, such a collation operation performed mechanically can be performed by a computer, and a certain specified attribute is selected, and a part of the item having a similar or identical attribute is selected by a full-text search using a computer. If a search is made and a judgment is made based on the item of each attribute and the related attribute, the range of the search can be narrowed, and the required data can be quickly searched.

By specifying the attribute in the "designating" step, related data can be easily found from the file recording segment. Taking FIG. 2 as an example, if the attribute specified by the user is “Taichung”, the file number A00
Two data, 0025 and A0000035, can be found (see FIG. 4). The related attribute number of the two data is 005. Therefore, the search is quickly performed according to the user's specification.

When the "designation" step is completed, the system advances the attribute and data retrieval steps. For example, suppose the attribute specified by the user is "America". Its attribute number is 0003. Next, in the "retrieving related attribute" step, "Canada" which is the attribute of the related attribute number 0003 is found from the attribute recording segment, and in the "step of retrieving data related to the designated attribute", the related attribute is retrieved from the file recording segment. Find the file with the number 003. Finally, in the "step of searching for data related to the related attribute of the designated attribute", data having a related attribute number of 007 (Canada) is found from the file recording segment (see FIG. 5). Through the above steps, the user can accurately and quickly search for the attribute data by designating it, and at the same time, detect other data related to the data and provide it to the user.

When there are many attributes specified by the user, the range of data search can be narrowed by performing processing such as cross concentration or connection concentration, and more accurate search can be achieved. For example, 199
When retrieving data on the overview of Taiwan-Canada trade in fiscal year 0, the user can easily specify four attributes: "1990", "Taiwan", "Canada", and "trade". Suppose there are 20 data with the attribute of "Taiwan"
If you have 15 data with the attribute "Canada"
By performing the cross-concentration logic operation on data having both the attributes “Taiwan” and “Canada”, the data is probably 5
If you specify the attribute of "trade", it will probably be narrowed down to two data. Thus, the required data can be more accurately searched by the simple calculation of the alley and sky.

[0028]

As is clear from the above description of the specific steps and the embodiment, the management of the database according to the present invention,
The search method has the following advantages.

1. By actively defining attributes, there is no restriction on fixed columns. A feature of the present invention is that all the attribute items of the document data to be defined are actively structured starting from the user's point of view. Such an active attribute definition method not only solves the problem of insufficient definition of document data caused by the influence of fixed columns, but also allows the user to subjectively define the characteristics of document data. When a user performs a search, the attribute can be accurately specified, and required data can be quickly searched. In other words, the present invention can be applied to management of data across different fields.

2. Applicable to different types of data formats. The so-called information includes data expressed in a video or audio format in addition to a text format. For these types of data, the search cannot be advanced by the full-text search method. However, according to the present invention, the non-character data is inductively and rearranged in a form in which a system is constructed from the outside. It is also useful for searching. In addition, because it is a form to build a system from the outside. No problem occurs when searching and retrieving without affecting the integrity of the original data.

3. It is possible to support a commonly used existing classification method. Since the present invention defines the attributes of data in the form of constructing a system from the outside, it can support a commonly used classification system and can classify one data into a plurality of different classification systems at the same time. And there is no need to duplicate the data. For example, users are at the bottom of the taxonomy, the US National Book Classification Standard, and the Republic of China (Taiwan)
When two attribute items of the National Book Classification Standard are constructed and the two classification systems are classified, any book is classified into any of the two classification systems. The formation of such categories is based on the definition of attributes. Therefore, the present invention is compatible with a plurality of classification systems and can be freely extended and modified.

4. Has ideal compatibility. When data attributes are used, they can be generally classified into two different types of attributes. One is the keyword attribute,
Represents people, things, times, places, things, etc. Thus, the word itself has independent and apparent significance. Another type of attribute is representative of the significance of the classification in the classification system. If these two types of attributes can be used simultaneously, a complete and appropriate definition can be made for the data. However, these two types of attributes have a conflicting relationship in terms of dispersion and structuring. According to the present invention, by actively defining attributes, it is possible to eliminate conflicts in the distribution and structuring of the two types of attributes and make full use of them.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram describing an example of an attribute recording segment according to the present invention.

FIG. 2 is an explanatory diagram describing an example of a file recording segment according to the present invention.

FIG. 3 is an explanatory diagram showing an example in which file recording is added in the present invention.

FIG. 4 is an explanatory diagram showing a record of a file to be searched in the present invention.

FIG. 5 is an explanatory diagram showing a record of a file to be searched in the present invention.

[Explanation of symbols]

Claims (9)

[Claims] 1. A method for managing and searching a database, which comprises a management part and a search part and actively defines data attributes, wherein the management part starts from a user's point of view. Defining an attribute for the attribute, searching for a related attribute related to the data from the defined attribute to form a dependency, and storing the data in a specific format based on the attribute and the related attribute. And storing the data in a specific location and displaying the attribute number associated with the location of the file, respectively, and associating with the defined attribute, wherein the portion of the search comprises: “Specified process” where the user specifies the attribute, “Related attribute search process” where the attribute of the specified condition is detected from the attribute record segment, and the attribute specified from the file record segment. Searching for document data related to the specified attribute, and searching for document data related to the specified attribute related to the specified attribute from the file recording segment. Searching for data related to the attribute ". 2. The “attribute definition” defines an attribute for data to be stored and directed from a user's own viewpoint, and is restricted by a fixed column or a normal form. 2. The database management and search method according to claim 1, wherein there is no database. 3. The step of defining a related attribute is a step of finding a related attribute related to a newly defined attribute from existing attribute items and forming a dependent relationship between the attribute and the attribute. 2. The database management and search method according to claim 1, wherein: 4. The method according to claim 1, wherein the step of storing the attribute includes storing, in a specific format, the attribute items for which the attribute recording segment has been constructed and whose definition has been completed. The “name”, “attribute number”, and “related attribute number” are recorded in the attribute record segment, the “attribute name” is actively input by the administrator, and the “attribute number” is the record of the attribute record segment. A number automatically generated each time an attribute is recorded by the management unit, wherein the “related attribute number” is for the user to find an attribute item related to the record from existing attribute items. The database management and search method according to claim 1. 5. The “save file” is executed by constructing a file recording segment, and the file recording segment includes a “file number”, a “storage location”, and the storage. The database management and search method according to claim 1, wherein an "attribute number" of the data file is recorded. 6. The "storage location" indicates a location at which a file is to be stored, and when executed by a computer, indicates the name of a disk unit and the name of a root and a file. The database management and search method according to claim 5, characterized in that: 7. The database management and search method according to claim 4, wherein the “attribute number” is displayed as a consecutive number in the order before and after the storage. 8. The database management and search method according to claim 1, wherein, in the step of specifying the attribute in the search part, the process is performed in a browsing or one-by-one matching manner. 9. The database management and search method according to claim 1, wherein, in the attribute designation step in the search part, a search range is narrowed down by a logic operation.