JP2008015774A - Counterfeit document detection system and program - Google Patents

Abstract

An object of the present invention is to provide a counterfeit document detection system and program capable of detecting similar parts between a plurality of documents.
In a counterfeit document detection system, a combination of an input interface unit for inputting a plurality of document files, a document storage unit for storing the input document files, and an input document file The imitation inspection unit 4 that inspects the imitation part and outputs the inspection result, and sequentially extracts a set of document files (document file 7) from the document file group stored in the document storage unit 3, and the imitation The imitation inspection driving unit 5 that holds the inspection result input to the inspection unit 4 and output from the counterfeit inspection unit 4 for the set of document files, and the inspection result held by the imitation inspection driving unit 5 And a result display unit 6 for presenting the imitation relationship between the stored document files 7.
[Selection] Figure 1

Description

  The present invention relates to a counterfeit document detection system and program capable of detecting a counterfeit portion between a plurality of documents.

In recent years, the digitization of report creation (use of a word processor, etc.) has become common, and cases of plagiarizing reports of others (for example, seniors) and Web pages on the Internet are increasing. The burden on teachers to check this is heavy, and there is a possibility of oversight, so computer support is necessary. Especially in e-Learning, since reports and the like are submitted in an electronic format, and the behavior of the students cannot be seen directly, it is meaningful to check imitation electronically.
JP 2003-30238 A

  However, heretofore, there has not been a system for detecting imitation parts that are similar to each other among a plurality of documents to be searched.

  As a related well-known technique, there is a technique (for example, Patent Document 1) for searching a document corresponding to a search key from a set of documents. By using the technique, a related document that is likely to be similar is searched. However, when a report imitation check is performed, the teacher eventually has to check the report to be checked against the related documents one by one. Against this background, there was a keen desire to automate check operations that relied on human hands.

  In view of the above problems, an object of the present invention is to provide a counterfeit document detection system and program capable of detecting a similar portion between a plurality of documents.

  In the imitation document detection system according to claim 1 of the present invention, an input interface unit for inputting a plurality of document files, a document storage unit for storing the input document files, and a set of input document files The imitation inspection unit that inspects the imitation part and outputs the inspection result, and sequentially extracts a set of document files from the document file group stored in the document storage unit, and inputs the document file to the imitation inspection unit. The imitation inspection driving unit that holds the inspection result output by the counterfeit inspection unit for the set of, and the imitation relationship between the accumulated document files based on the inspection result that the imitation inspection driving unit holds And a result display unit for presenting.

  In this way, it is possible to detect imitation documents by inspecting imitation between a plurality of document files input to the system.

  In the counterfeit document detection system according to claim 2 of the present invention, the document storage unit can classify the document file to be stored, and the imitation inspection driving unit performs classification in the document storage unit and an instruction from the user. The document file sets are sequentially extracted from the document storage unit and input to the counterfeit inspection unit, and the inspection result output by the counterfeit inspection unit is held for the document file set. It is characterized by that.

  In this way, the user can freely specify the classification to be subjected to the imitation inspection, and can perform the imitation inspection on the document file belonging to the classification.

  In the counterfeit document detection system according to the third aspect of the present invention, the result display unit displays the document file name of the set of document files having a counterfeit relationship and the degree of imitation for the set of document files in a tabular form. In addition, the imitation part is displayed for an arbitrary set of document files in the list of sets of document files.

  In this way, it is possible to confirm at a glance a document file having a counterfeit relationship from the list, and it is possible to specifically confirm which part in the document file is the imitation part.

  In the imitation document detection system according to claim 4 of the present invention, an input interface unit for inputting a document file, a document storage unit for storing the document file input via the document input interface unit, and an input A search word group is generated based on the text described in the document file, and an electronically accessible document outside the system is searched based on the generated search word group, and a document group as a search result is obtained. A search result document obtained by assigning an identifier for identifying a document group of the search result obtained by inputting the input document file to the document search unit and outputting the document group of the search result. A search result document accumulating unit that accumulates as a file, an imitation inspection unit that inspects an imitation part against a set of input document files, and outputs the inspection result; The search result document file is sequentially extracted from the search result document storage unit, and a set is made with the document file that has been input and stored in the document storage unit, and the set of document files is input to the imitation checking unit, The imitation inspection driving unit that holds the inspection result output by the counterfeit inspection unit for the set of document files, and the document file that has been input based on the inspection result that the imitation inspection driving unit holds A result display unit that presents the imitation relationship with the search result document file searched by the document search unit.

  In this way, whether the input document file is a counterfeit document by checking imitation between the document file to be inspected input to the system and a large number of unspecified related documents retrieved from outside the system. Whether or not can be determined in a wider document range.

  In the counterfeit document detection system according to claim 5 of the present invention, the result display unit includes a document file identifier of the search result document file having a counterfeit relationship with the input document file, and a degree of imitation of the search result document file. Are displayed as a list in a table format, and for any of the search result document files in the document file identifier list, an imitation part with the input document file is displayed.

  In this way, it is possible to confirm at a glance a document file having a counterfeit relationship from the list, and it is possible to specifically confirm which part in the document file is the imitation part.

  In the counterfeit document detection system according to claim 6 of the present invention, the input interface unit for inputting a plurality of document files, the document storage unit for storing the input document files, and the input document file are described. A document search unit that generates a search word group based on the generated text, searches an externally electronically accessible document based on the generated search word group, and outputs a document group of the search result; For each of the document files stored in the document storage unit, an identifier capable of identifying the document group of the search result obtained by inputting the document file to the document search unit is assigned to each document file. A search result document storage unit that stores the result as a search result document file, and a counterfeit inspection unit that inspects the imitation part against the set of input document files and outputs the inspection result; A set of document files is sequentially extracted from the document file group stored in the document storage unit and the search result document file group stored in the search result document storage unit, and is input to the imitation checking unit, and the document A counterfeit inspection driving unit that holds the inspection result output by the counterfeit inspection unit for a set of files, and between the stored document files and input based on the inspection result held by the counterfeit inspection driving unit And a result display unit for presenting a mimicry relationship between the retrieved document file and the retrieval result document file retrieved by the document retrieval unit.

  In this way, the imitation between a plurality of document files input to the system is checked, and the imitation between the document file input to the system and an unspecified number of related documents searched from outside the system is checked. Thus, it can be determined in a wider document range whether the input document file is a counterfeit document.

  In the imitation document detection system according to claim 7 of the present invention, the document storage unit and the search result document storage unit can classify the document file and the search result document file to be stored, and the imitation inspection driving unit Based on the classification in the document storage unit and the search result document storage unit and the instruction from the user, the set of document files is sequentially extracted from the document storage unit and the search result document storage unit, and the imitation inspection unit The inspection result output by the counterfeit inspection unit for the set of document files is held.

  In this way, the user can freely specify the classification to be subjected to the imitation inspection, and can perform the imitation inspection on the document file belonging to the classification.

  In the counterfeit document detection system according to claim 8 of the present invention, the result display unit displays a list of document file names or document file identifiers of the set of document files having a counterfeit relationship, and imitation levels for the set of document files. In addition to displaying in tabular form, the imitation part is displayed for an arbitrary set of document files in the list of sets of document files.

  In this way, it is possible to confirm at a glance a document file having a counterfeit relationship from the list, and it is possible to specifically confirm which part in the document file is the imitation part.

  In the counterfeit document detection system according to claim 9 of the present invention, the result display unit uses a set of document files in the set as a node for the set of sets of document files having a counterfeit relationship, and imitates between the document files. When there is a relationship, the nodes of the document file having the imitation relationship are connected by a branch, the visual characteristics of the branch are determined based on the degree of imitation between the document files, and displayed as a graph And

  In this way, the inspection result of the document file is graphed, so that the imitation relationship between the document files becomes visually clear, and the user can easily understand the inspection result.

  According to a program of a tenth aspect of the present invention, a computer is caused to function as the counterfeit document detection system according to any one of the first to ninth aspects.

  In this way, a counterfeit document detection system can be easily constructed by a computer.

  According to the first aspect of the present invention, it is possible to provide a counterfeit document detection system capable of detecting a similar portion between a plurality of documents.

  According to claim 2 of the present invention, the inspection efficiency can be improved by making it possible to select the imitation inspection object by classification.

  According to the third aspect of the present invention, the imitation inspection result can be presented to the user in an easily understandable manner.

  According to claim 4 of the present invention, it is possible to provide a counterfeit document detection system capable of detecting imitation in a wider document range using related documents outside the system.

  According to claim 5 of the present invention, the imitation inspection result can be presented to the user in an easily understandable manner.

  According to claim 6 of the present invention, it is possible to provide a counterfeit document detection system that can detect a similar portion between a plurality of documents and can detect imitation in a wider document range by using a related document outside the system. .

  According to claim 7 of the present invention, the inspection efficiency can be improved by making it possible to select the imitation inspection object by classification.

  According to claim 8 of the present invention, the imitation inspection result can be presented to the user in an easily understandable manner.

  According to the ninth aspect of the present invention, it is possible to visually present the imitation inspection result to the user in an easy-to-understand manner.

  According to claim 10 of the present invention, a counterfeit document detection system can be easily constructed by a computer.

  Hereinafter, preferred embodiments of a counterfeit document detection system and program according to the present invention will be described with reference to the accompanying drawings. In addition, in each Example, the same code | symbol is attached | subjected to the same location, Since description of a common part overlaps, it abbreviate | omits as much as possible. In each embodiment, only the counterfeit document detection system will be described, but the counterfeit document detection program of the present invention causes the installed computer to function in the same manner as each component of the system of each embodiment. The description is omitted.

  FIG. 1 is a system configuration diagram of a counterfeit document detection system in the first embodiment. The counterfeit document detection system 1 includes an input interface unit 2, a document storage unit 3, a counterfeit inspection unit 4, a counterfeit inspection drive unit 5, and a result display unit 6.

  The document input interface unit 2 allows a user to input a plurality of electronic format files 7 (hereinafter referred to as document files 7) to the document storage unit 3 using operation input means such as a keyboard and a mouse. An operation screen or the like that is actually operated by the user corresponds to this. The document file 7 may be any document in any field, such as a report, a paper, a novel, or a newspaper article.

  The configuration of the document input interface unit 2 is slightly different depending on the system configuration of the counterfeit document detection system 1. That is, if the counterfeit document detection system 1 is a system constructed by using a network such as the Internet or a LAN, for example, the document input interface unit 2 corresponds to a client operated by the user, and is transmitted to the server via the network. Data communication means for uploading the document file 7 to the built-in document storage unit 3 is provided. In the client, the operation screen is provided to the user by, for example, a dedicated application or a Web page.

  The document storage unit 3 receives and stores the input document file 7 from the document input interface unit 2 and corresponds to, for example, a database stored in a storage device such as a hard disk. The document storage unit 3 stores and manages the document file 7 in the system. In particular, the document storage unit 3 is configured to be able to classify the document file 7 to be stored based on an instruction from the user, for example, every report submission year when the document file 7 is stored. The classification is not limited to an instruction from the user, and may be automatically performed based on a preset classification standard.

  The counterfeit inspection unit 4 inspects the imitation part of the two input document files 7 and outputs the imitation part and the imitation degree as a result. The imitation inspection algorithm will be described later. .

  The counterfeit inspection driving unit 5 sequentially takes out a set of two document files 7 from the document file group stored in the document storage unit 3 based on the classification in the document storage unit 3 and the instruction from the user, and the counterfeit inspection unit 4. And the inspection result output by the counterfeit inspection unit 4 for the set of the document files 7 is held.

  If the counterfeit document detection system 1 is a system constructed using a network, the document storage unit 3, the counterfeit inspection unit 4, and the counterfeit inspection drive unit 5 are provided on the server side in response to a request from the client. The counterfeit inspection process is executed, and only the result of the counterfeit inspection process is returned to the requesting client.

  The result display unit 6 presents the imitation relationship between the accumulated document files 7 to the user based on the imitation inspection result held by the imitation inspection driving unit 5, and the user actually visually recognizes the result. An inspection result display screen or the like to correspond to this. More specifically, the result display unit 6 receives the inspection result from the counterfeit inspection driving unit 5 and lists the document file name of the set of document files 7 having the imitation relationship and the imitation degree for the set of document files 7. Are displayed in a tabular form, and an imitation portion is displayed for an arbitrary set of document files 7 in the list of sets of document files 7 by a user operation. Further, as another display mode of the inspection result, with respect to a set of sets of document files 7 having imitation relation, each document file 7 included in the set is a node, and there is a imitation relation between the document files 7, The nodes of the document file 7 having the imitation relation are connected by branches, and the visual characteristics such as length, thickness, color, etc. of the branches are determined based on the degree of imitation between the document files 7, and a graph is obtained. indicate.

  Similar to the document input interface unit 2, the result display unit 6 corresponds to the client operated by the user if the counterfeit document detection system 1 is a system constructed using a network. The data communication means for downloading the inspection result data from the counterfeit inspection driving unit 5 built in the server is provided. In the client, the inspection result display screen or the like is provided to the user by, for example, a dedicated application or a Web page.

  Here, the specific Example of the imitation inspection algorithm mounted in the imitation inspection part 4 is demonstrated. The counterfeit inspection unit 4 uses a similar character analysis method to detect a similar character string existing between documents. In addition to these, for example, a well-known similar character analysis method such as a vector space method can be applied. Since each of these analysis methods has advantages and disadvantages, the user may be able to select at the time of imitation inspection.

  First, n-gram analysis (character n-gram analysis) will be described. This is the most basic n-gram analysis, and matches are detected using n-grams in character units. Hereinafter, the character n-gram analysis will be described with a specific example. For example, it is assumed that there is a document A “Ai Ueoki Kakusashi Ai Ueokiki” and a document B “Ueokaki Kaketsutsutsu”. The 5-gram of each document is as shown in FIG. The parenthesis is the position of the first character in the n-gram. When a pair that matches each other is found, A (03) -B (01), A (04) -B (02), A (05) -B (03), A (16) -B (01 ) 4 sets.

  Thus, the matrix of coincidence detection results is as shown in FIG. In the figure, document A is in the horizontal direction and document B is in the vertical direction, and the element corresponding to the mutually matching set is 1. In the figure, 0 is actually entered as a value at a place other than 1, but it is omitted here. An n-gram corresponding to a place where 1 is entered is basically detected as a similar part, that is, an imitation part, but an imitation part of an n-gram set corresponding to a part where 1 is continuously arranged diagonally. You may detect as. The coincidence detection can be performed by hitting each entry of the n-gram of the document A and each entry of the n-gram of the document B.

  However, a more efficient processing method can be considered. Each entry of n-gram is sorted in dictionary order, and the position information of the same entry is given to the entry as a list. That is, the list shown in FIG. 6 is called an inverted list. In this example, the n-gram entries were mostly in lexicographic order (the original document was the content), so it seems that there is not much difference, but in the case of an actual document, It should look very different from the n-gram list.

  Since the entries in the inverted list are sorted in lexicographic order, it is not necessary to make a brute force check for matching entries between document A and document B, using the pointers for document A and B, A match entry can be detected by comparing the entries at the positions indicated by both pointers and advancing the pointer indicating the smaller one. Assuming that the number of entries in document A is N and the number of entries in document B is M, a total of N × M comparisons is required. On the other hand, if the number of entries in the inverted list is N ′ and M ′, respectively, the number of comparisons using N ′ and M ′ is sufficient, so that the number of comparisons (= processing amount) can be greatly reduced. The result is of course the same).

  Next, word n-gram analysis will be described as another n-gram analysis. In this analysis method, words (independent words obtained by morphological analysis) are used instead of characters. If Document A is "Horyuji, the bell rings if you eat it," and Document B is "The next customer is a customer who eats well," a list of free-standing words obtained by morphological analysis is (by the morphological analyzer) Document A becomes <柿> <eat> <bell> <ring> <Horyuji> and document B becomes <neighbor> <customer> <good> <柿> <eat> <customer>. Taking this word 2-gram, it is as shown in FIG. When matching is examined in the figure, A (01) -B (04) is obtained. The matrix of the coincidence detection result is the same as in the case of the character n-gram. The same applies to the coincidence detection algorithm.

  Hereinafter, the operation of the counterfeit document detection system 1 constructed by the above configuration will be described together with its operation method. Here, a case where a client server system using a network is adopted as the counterfeit document detection system 1 and a counterfeit inspection is performed on a report as the document file 7 is considered.

  The user prepares a report to be analyzed, and accesses an operation screen page provided by the document input interface unit 2 using a client Web browser or the like. A report is uploaded to the document storage unit 3 of the server through the operation screen page, and an analysis execution operation is performed. At this time, the selection of the analysis comparison target classification, the selection of the analysis method, and the like are also performed.

  Upon receiving the analysis execution request from the client, in the server, the counterfeit inspection driving unit 5 sequentially extracts a set of two reports to be compared from among a plurality of report groups stored in the document storage unit 3, and the counterfeit inspection unit Input to 4. The imitation inspection unit 4 inspects imitation parts that are similar to each other between the two input reports according to the above-described imitation inspection algorithm, and imitates the detected imitation part and its imitation degree (similarity) as an inspection result. Output to the inspection drive unit 5. The inspection result is transmitted from the counterfeit inspection driving unit 5 to the result display unit 6, and embedded in the Web page by the result display unit 6, whereby the analysis result display page is displayed on the client side.

  FIG. 8 shows the analysis result display page 10. At the bottom of the page, a list 11 of a set of two reports in which imitation detection is performed is displayed. In this list 11, from the left, the file name of one report (file name 1), the file name of another report (file name 2), the longest match count, the total match count, and the link 16 to the detailed display (See details). These displays may be appropriately changed according to the analysis result item. When the analysis method link 12 at the top is clicked, the display content of the list 11 is switched to the result list by the analysis method, and the list can be displayed by analysis method. The analysis result to be displayed in the list 11 can be specified by the list box 13. The display type includes "longest match number, overall match number, original frequency of file 1". You can specify the “number, number, all items” and manipulate the results displayed in the list 11. The designation is confirmed by clicking the redisplay button 14, and the display content of the list 11 is changed and redisplayed. The original frequency of each report is calculated according to the calculation formula (1- (number of matched grams / total number of grams)) × 100. The calculated original frequency is written in red after the file names listed in the list 11, for example. Here, the calculation formula for the original power is merely an example in the present embodiment, and the original power in the present invention is not limited to the calculation formula shown here.

  When the link 16 is clicked, a detailed display page shown in FIG. 9 is displayed. In this page, the contents of the two document files 7 corresponding to the set of file name 1 and file name 2 in the list 11 are displayed in the left and right divided areas, respectively. The matched part (imitation part) is highlighted in color. Numbers that match are numbered, and the same numbers on the left and right correspond to each other. Clicking on the number will link to the corresponding location in the page. The longest match is displayed in bold.

  When the graph display button 15 labeled “Grouped and display image” in FIG. 8 is clicked, a graph display page as shown in FIG. 10 is displayed. This shows the correlation in the list 11 displayed in FIG. The correlation graph 20 shows a set of document files 7 having a mimicry relationship, and each document file 7 included in the set has a node as a node. The nodes of the file 7 are connected by branches, and visual features such as length, thickness, color, etc. of the branches are determined based on the degree of imitation between the document files 7 and displayed as a graph. Symbols and numerical values displayed in the correlation graph 20 include, for example, a dark circle as a node: a file in this folder, a thin circle as a node: a file in a folder added to the comparison target, a line as a branch: for each file It may be drawn as a connection, a red character: a file number, a light blue character: the total number of matches, etc., which are appropriately changed according to a parameter indicating a correlation between the document files 7 or a design convenience.

  As described above, in the counterfeit document detection system 1 according to the first embodiment, the input interface unit 2 for inputting a plurality of document files 7, the document storage unit 3 for storing the input document files 7, and the input The imitation checking unit 4 that inspects the imitation part of the set of document files and outputs the inspection result, and the document file group (document file 7) sequentially from the document file group stored in the document storage unit 3 The imitation inspection driving unit 5 that holds the inspection result output from the imitation inspection unit 4 for the set of document files and is output to the imitation inspection unit 4 and the imitation inspection driving unit 5 holds And a result display unit 6 for presenting the imitation relationship between the stored document files 7 based on the inspection result.

  In this way, it is possible to detect imitation documents by inspecting imitation between a plurality of document files 7 input to the system. Therefore, it is possible to provide a counterfeit document detection system that can detect similar portions between a plurality of documents.

  Further, in the counterfeit document detection system 1 according to the first embodiment, the document storage unit 3 can classify the document file 7 to be stored, and the counterfeit inspection driving unit 5 performs classification in the document storage unit 3. Based on the instructions from the user, the document file sets are sequentially taken out from the document storage unit 3 and input to the imitation checking unit 4, and the imitation checking unit 4 outputs the document file sets. It is characterized by holding inspection results.

  In this way, the user can freely specify the classification to be subjected to the imitation inspection, and can perform the imitation inspection on the document file 7 belonging to the classification. Therefore, the inspection efficiency can be improved by making it possible to select the imitation inspection target by classification.

  Furthermore, in the counterfeit document detection system 1 according to the first embodiment, the result display unit 6 uses a tabular format as a list of document file names of the document file sets having a counterfeit relationship and imitation levels for the document file sets. And an imitation part is displayed for an arbitrary set of document files in the list of sets of document files.

  In this way, the document file 7 having the imitation relationship can be confirmed at a glance from the list, and it can be specifically confirmed which portion in the document file 7 is the imitation portion. Therefore, the imitation inspection result can be presented to the user in an easy-to-understand manner.

  Further, in the counterfeit document detection system 1 of the first embodiment, the result display unit 6 uses the individual document files included in the set as a node for the set of sets of document files having a counterfeit relationship, When there is an imitation relationship, the nodes of the document files having the imitation relationship are connected by branches, the visual characteristics of the branches are determined based on the degree of imitation between the document files, and displayed as a graph It is characterized by.

  In this way, the inspection result of the document file is graphed, so that the imitation relationship between the document files becomes visually clear, and the user can easily understand the inspection result. Therefore, the imitation inspection result can be visually presented to the user in an easy-to-understand manner.

  The first embodiment can also be realized by a program for causing a computer to function as the counterfeit document detection system 1.

  In this way, a counterfeit document detection system can be easily constructed by a computer.

  FIG. 2 is a system configuration diagram of the counterfeit document detection system in the second embodiment. The counterfeit document detection system 51 includes an input interface unit 2, a document storage unit 3, a document search unit 57, a search result document storage unit 59, a counterfeit inspection unit 4, a counterfeit inspection drive unit 55, and a result display unit 56. And is configured. The input interface unit 2, the document storage unit 3, and the counterfeit inspection unit 4 have substantially the same configuration as in the first embodiment, but in the second embodiment, one document file 7 is input to the document input interface unit 2. It has become. This is because, in the counterfeit document detection system 51 in the second embodiment, for example, an unspecified number of system external document files 57 as search result document files searched using a search engine such as the Internet, and the inspection target. This is because imitation inspection is performed with the document file 7. Of course, this does not limit the number of documents that can be input from the document input interface unit 2 in this patent to one document. When a plurality of documents are input, after the plurality of documents are stored in the document storage unit 3, the operations described in this embodiment may be repeated while taking out the documents from the document storage unit 3 one by one. .

  Based on the text described in the input document file 7, the document search unit 57 searches the system external document file 58 that is an electronically accessible document outside the system that is related to the document in terms of content. The resulting document group is output. As the system external document file 58, for example, a Web page or an electronic document file published on the Internet corresponds. Strictly speaking, the Web page or the like is stored as a system external document file 58 after an appropriate identifier is given in a search result document storage unit 59 described later. The document search unit 57 can be realized in various configurations by combining existing technologies. For example, a group of web pages on the Internet is collected across the web by following the link information described in the web page, and the similarity between the collected individual pages and the input document file 7 is determined between the two documents. The web page having a high degree of similarity is output as a result, and the word group that is the subject of the document based on the appearance frequency from the text of the input document file 7 And a search result output from the search engine is output as a search result of the document search unit 57 using the Web document full-text search service provided on the Internet as a search term. The structure etc. to perform are considered.

  The search result document storage unit 59 stores the system external document file 58 as a document group of the search results obtained by the document search unit 57 with an identifiable identifier assigned to each individual document. For example, a database stored in a storage device such as a hard disk corresponds to this. The search result document storage unit 59 saves and manages the system external document file 58 in the system. In particular, the search result document storage unit 59 is configured to be able to classify the system external document file 58 to be stored based on an instruction from the user when the system external document file 58 is stored.

  The counterfeit inspection driving unit 55 sequentially extracts the system external document files 58 one by one from the search result document storage unit 59, creates a pair with the document file 7 that has been input and stored in the document storage unit 3, and the document file (A document file 7 and a system external document file 58) are input to the counterfeit inspection unit 4, and the inspection result output by the counterfeit inspection unit 4 for the set of document files is held.

  The result display unit 56 presents the imitation relationship between the accumulated document file 7 and the system external document file 58 to the user based on the imitation inspection result held by the imitation inspection driving unit 5. An inspection result display screen or the like actually viewed by the user corresponds to this. More specifically, the result display unit 6 displays the document file identifier of the system external document file 58 having the imitation relationship with respect to the input document file 7 and the imitation degree for the system external document file 58 as a list in a table format. At the same time, the imitation part with the document file 7 input for any system external document file 58 in the document file identifier list by the user's operation is displayed. Further, as another display mode of the inspection result, for a set of the document file 7 having imitation relation and the system external document file 58, the individual document file 7 and the system external document file 58 included in the set are nodes. When there is a mimicry relationship between the document file 7 and the system external document file 58, the nodes of the document file 7 and the system external document file 58 having the mimicry relationship are connected by branches, and the document file 7 and the system external document file are connected. Based on the degree of imitation between 58, visual features such as length, thickness and color of the branch are determined and displayed as a graph.

  As described above, in the counterfeit document detection system 51 of the second embodiment, the input interface unit 2 for inputting the document file 7 and the document file 7 input via the document input interface unit 2 are stored. A search term group is generated based on the text described in the document storage unit 3 and the input document file 7, and an electronically accessible document outside the system is generated based on the generated search term group. A document search unit 57 for searching and outputting a document group of the search result, and the document group of the search result obtained by inputting the input document file 7 to the document search unit 57 for each document A search result document storage unit 59 that stores an identifier that can be identified and stored as a system external document file 58 corresponding to the search result document file, and a set of input document files. The imitation inspection unit 4 that inspects the copied portion and outputs the inspection result, and the system external document file 58 is sequentially taken out from the search result document storage unit 59, and is input and stored in the document storage unit 3 A set with the document file 7 is created, and the set of the document file (document file 7, system external document file 58) is input to the imitation checking unit 4, and the imitation checking unit 4 outputs the set of the document files. The imitation inspection driving unit 55 for holding the inspection result to be performed, and the system searched by the input document file 7 and the document searching unit 57 based on the inspection result held by the imitation inspection driving unit 55 And a result display unit 56 for presenting the imitation relationship with the external document file 58.

  In this way, by checking imitation between the document file 7 to be inspected input to the system and a large number of unspecified related documents retrieved from outside the system, the input document file 7 is a counterfeit document. It is possible to determine whether or not there is a wider document range. Therefore, it is possible to provide a counterfeit document detection system capable of detecting imitation in a wider document range using related documents outside the system.

  In the counterfeit document detection system 51 of the second embodiment, the result display unit 56 includes the document file identifier of the system external document file 58 having a counterfeit relationship with the input document file 7, and the system external document file. The imitation level for 58 is displayed as a list in a table format, and the imitation part of the input document file 7 is displayed for any of the system external document files 58 in the document file identifier list. It is characterized by.

  In this way, it is possible to confirm at a glance a document file having a counterfeit relationship from the list, and it is possible to specifically confirm which part in the document file is the imitation part. Therefore, the imitation inspection result can be presented to the user in an easy-to-understand manner.

  FIG. 3 is a system configuration diagram of the counterfeit document detection system in the third embodiment. In the third embodiment, a system in which the first embodiment and the second embodiment are combined is provided. That is, the counterfeit document detection system 61 includes an input interface unit 2, a document storage unit 3, a document search unit 57, a search result document storage unit 59, a counterfeit inspection unit 4, a counterfeit inspection drive unit 65, and a result display. And 66. The input interface unit 2, the document storage unit 3, and the counterfeit inspection unit 4 have substantially the same configurations as those of the first embodiment, and the document search unit 57 and the search result document storage unit 59 have substantially the same configurations as those of the second embodiment. .

  The counterfeit inspection driving unit 65 sequentially selects two documents from the document file group (document file 7) stored in the document storage unit 3 and the document group (system external document file 58) stored in the search result document storage unit 59. A set of files (a set of the document file 7 or a set of the document file 7 and the system external document file 58) is taken out and input to the imitation checking unit 4, and the imitation checking unit 4 outputs the set of document files. The inspection result is held.

  The result display unit 66 determines the imitation relationship between the stored document files 7 and between the document file 7 and the system external document file 58 based on the imitation inspection result held by the imitation inspection driving unit 5. The inspection result display screen or the like that the user actually visually recognizes corresponds to this. More specifically, the result display unit 6 displays the document file name or document file identifier of the set of document files in the imitation relationship and the imitation level for the set of document files in a tabular form as a list, As a result of the above operation, an imitation portion is displayed for an arbitrary document file set in the list of document file sets. Further, as another display mode of the inspection result, with respect to a set of the document file set having the imitation relationship, the individual document file 7 and the system external document file 58 included in the set are set as nodes, and the set of the document file is set. When there is an imitation relationship between the nodes of the document file 7 and the system external document file 58 having the imitation relationship, the nodes are connected by a branch, and the length of the branch is determined based on the imitation degree between the document file sets. Visual characteristics such as thickness, thickness, and color are determined and displayed as a graph.

  As described above, in the counterfeit document detection system 61 of the third embodiment, the input interface unit 2 for inputting a plurality of document files 7, the document storage unit 3 for storing the input document files 7, and the input A search word group is generated based on the written text described in the document file 7, and an external electronically accessible document is searched based on the generated search word group, and the search result document A document search unit 57 for outputting a group, and the search result document obtained by inputting the document file 7 to the document search unit 57 for each document file 7 stored in the document storage unit 3 A search result document storage unit 59 for storing a group as a system external document file 58 corresponding to a search result document file after assigning an identifiable identifier to each document, and a set of input document files The imitation inspection unit 4 that inspects the imitation part and outputs the inspection result, the document file group stored in the document storage unit 3, and the search result document file group stored in the search result document storage unit 59 A set of document files (document file 7, system external document file 58) is sequentially extracted from the list, input to the imitation inspection unit 4, and the inspection result output by the imitation inspection unit 4 for the set of document files Based on the inspection result held by the counterfeit inspection driving unit 65 and between the stored document files 7 and the input document file 7 and the document search unit 57 And a result display unit 66 for presenting the imitation relationship with the system external document file 58 retrieved by the above.

  In this manner, imitation between a plurality of document files input to the system is inspected, and imitation between the document file 7 input to the system and an unspecified number of related documents retrieved from outside the system is inspected. Thus, it is possible to determine whether or not the input document file 7 is a counterfeit document in a wider document range. Therefore, it is possible to provide a counterfeit document detection system capable of detecting a similar portion between a plurality of documents and detecting imitation in a wider document range by using a related document outside the system.

  In the counterfeit document detection system 61 of the third embodiment, the document storage unit 3 and the search result document storage unit 59 can classify the document file 7 and the system external document file 58 to be stored, The counterfeit inspection driving unit 65 sequentially sets the document file sets to the document storage unit 3 and the search result document storage based on the classification in the document storage unit 3 and the search result document storage unit 59 and an instruction from the user. It is extracted from the unit 59, input to the counterfeit inspection unit 4, and holds the inspection result output by the counterfeit inspection unit 4 for the set of document files.

  In this way, the user can freely specify the classification to be subjected to the imitation inspection, and can perform the imitation inspection on the document file belonging to the classification. Therefore, the inspection efficiency can be improved by making it possible to select the imitation inspection target by classification.

  Further, in the counterfeit document detection system 61 of the present embodiment, the result display unit 66 lists the document file name or document file identifier of the set of document files having a counterfeit relationship, and the imitation level for the set of document files. In addition to displaying in tabular form, the imitation part is displayed for an arbitrary set of document files in the list of sets of document files.

  In this way, it is possible to confirm at a glance a document file having a counterfeit relationship from the list, and it is possible to specifically confirm which part in the document file is the imitation part. Therefore, the imitation inspection result can be presented to the user in an easy-to-understand manner.

  In addition, this invention is not limited to the said Example, It can change in the range which does not deviate from the meaning of this invention.

It is a block diagram which shows the system configuration | structure of the imitation document detection system in 1st Example of this invention. It is a block diagram which shows the system configuration | structure of the imitation document detection system in 2nd Example of this invention. It is a block diagram which shows the system configuration | structure of the imitation document detection system in 3rd Example of this invention. It is explanatory drawing which shows the n-gram set produced by the character n-gram analysis about documents A and B. FIG. It is explanatory drawing which shows the matrix of the coincidence detection result in FIG. FIG. 5 is an explanatory diagram showing an inverted list in which the n-gram set in FIG. 4 is sorted in dictionary order. It is explanatory drawing which shows the n-gram set produced by the word n-gram analysis about documents A and B. FIG. It is explanatory drawing which shows the analysis result screen by the imitation document detection system in 1st Example of this invention. It is explanatory drawing which shows a detailed display screen of an analysis result same as the above. It is explanatory drawing which shows the correlation graph display screen of an analysis result same as the above.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Imitation document detection system 2 Input interface part 3 Document storage part 4 Imitation inspection part 5 Imitation inspection drive part 6 Result display part 7 Document file
51 Counterfeit document detection system
55 Imitation inspection drive
56 Result display area
57 Document Search Department
58 System external document file (Search result document file)
59 Search result document storage
61 Counterfeit document detection system
65 Imitation inspection drive
66 Result display area

Claims (10)

An input interface unit for inputting a plurality of document files, a document storage unit for storing the input document files,
A counterfeit inspection unit that inspects a counterfeit portion against a set of input document files and outputs the inspection result;
Imitation in which a set of document files is sequentially extracted from the document file group stored in the document storage unit, input to the imitation inspection unit, and the inspection result output by the imitation inspection unit for the set of document files is retained An inspection drive unit;
A counterfeit document detection system comprising: a result display unit that presents a counterfeit relationship between the stored document files based on the test result held by the counterfeit test drive unit. The document storage unit can classify the document files to be stored, and the imitation inspection driving unit sequentially sets the document file sets based on the classification in the document storage unit and instructions from the user. 2. The counterfeit document detection according to claim 1, wherein the counterfeit document detection unit retrieves from the storage unit, inputs to the counterfeit inspection unit, and holds the inspection result output by the counterfeit inspection unit for the set of document files. system. The result display unit displays the document file name of the set of document files having the imitation relationship and the imitation degree for the set of document files in a tabular form as a list, and an arbitrary one in the list of the set of document files The counterfeit document detection system according to claim 1 or 2, wherein the counterfeit portion is displayed for a set of document files. A submission interface for submitting document files;
A document storage unit for storing the document file input via the document input interface unit;
A search word group is generated based on the text described in the input document file, and an electronically accessible document outside the system is searched based on the generated search word group. A document search unit for outputting a document group;
A search result document that stores the search result document group obtained by inputting the input document file into the document search unit, with an identifier that can be identified for each document, and stored as a search result document file A storage unit;
A counterfeit inspection unit that inspects a counterfeit portion against a set of input document files and outputs the inspection result;
The retrieval result document file is sequentially taken out from the retrieval result document storage unit, and a pair is formed with the document file that is input and stored in the document storage unit, and the set of document files is input to the imitation inspection unit. , A counterfeit inspection driving unit that holds an inspection result output by the counterfeit inspection unit for the set of document files;
A result display unit that presents a mimicry relationship between the input document file and the search result document file searched by the document search unit based on the inspection result held by the counterfeit inspection drive unit; A counterfeit document detection system comprising: The result display unit displays the document file identifier of the search result document file having a counterfeit relationship with the input document file and the imitation degree for the search result document file in a tabular form, and the document file 5. The counterfeit document detection system according to claim 4, wherein a counterfeit part of the arbitrary search result document file in the identifier list with the input document file is displayed. A submission interface for submitting multiple document files;
A document storage unit for storing the input document file;
A search term group is generated based on the text described in the input document file, and an external electronically accessible document is searched based on the generated search term group, and the search result document A document search unit for outputting groups,
For each of the document files stored in the document storage unit, an identifier capable of identifying the document group of the search result obtained by inputting the document file to the document search unit is assigned to each document file. And a search result document storage unit for storing as a search result document file,
A counterfeit inspection unit that inspects a counterfeit portion against a set of input document files and outputs the inspection result;
A set of document files is sequentially extracted from the document file group stored in the document storage unit and the search result document file group stored in the search result document storage unit, and is input to the imitation checking unit, and the document A counterfeit inspection driving unit that holds the inspection result output by the counterfeit inspection unit for a set of files;
Imitation between the stored document files and between the input document file and the search result document file searched by the document search unit based on the inspection result held by the counterfeit inspection driving unit A counterfeit document detection system comprising a result display unit for presenting a relationship. The document storage unit and the search result document storage unit can classify the document file and the search result document file to be stored, and the imitation inspection driving unit includes the document storage unit and the search result document storage unit. The document file sets are sequentially extracted from the document storage unit and the search result document storage unit based on the classification and the instruction from the user and input to the imitation checking unit. The counterfeit document detection system according to claim 6, wherein the counterfeit inspection unit outputs an inspection result output by the counterfeit inspection unit. The result display unit displays the document file name or the document file identifier of the document file set having the imitation relationship and the imitation degree for the document file set as a list in a tabular form. 8. The counterfeit document detection system according to claim 6, wherein the counterfeit part is displayed for an arbitrary set of document files in the list. When the result display unit has a set of document files having a counterfeit relationship, each document file included in the set is a node, and when there is a counterfeit relationship between the document files, The node is connected by a branch, the visual feature of the branch is determined based on the degree of imitation between the document files, and displayed as a graph. The imitation document detection system described. A program for causing a computer to function as the imitation document detection system according to any one of claims 1 to 9.

Images