JP5395121B2 - Document revision management system and document revision management program - Google Patents

Description

  The present invention relates to a document revision management system and a document revision management program, and more particularly to processing when a document is revised.

  A document management system such as a document database generally has a function of accepting revisions of stored documents. Also, some document management systems have a function of identifying other documents that may be affected by a revision when a document is revised. An example of the configuration of such a document revision management system is shown in Patent Document 1. In items 2) and 3) on pages 15 to 16 of Patent Document 1, an e-mail is sent to the creator of another document so that the content of the revision is reflected in the other document according to the revision of the document. It is described to notify by.

International Publication No. 97/25679 Pamphlet

  However, the conventional technique has a problem in that it is not sufficiently known when the revision content of a document is not reflected in another related document. For example, in the configuration of Patent Document 1, an e-mail is transmitted to the creator of a related document in response to a revision of a certain document, but the revision is not notified to a general viewer who is not the creator. . For this reason, from the time a document is revised until the revision is reflected in the related document, the general viewer is not aware that the related document needs to be reflected in the state before the revision. May misunderstand that it is the latest content and use related documents.

  The present invention has been made to solve such a problem, and provides a document revision management system and a document revision management program capable of notifying the revision contents of a document when it is not reflected in other related documents. The purpose is to provide.

  In order to solve the above-described problems, a document revision management system according to the present invention includes a document storage unit that stores a plurality of documents in association with document IDs and document names, and a revision target selected from the documents. Revision instruction accepting section that accepts revision instructions for documents, and related documents specifying section that identifies documents related to revision instructions as related documents from among documents excluding revision target documents, and related documents Document revision information unit for creating revision source pointer including information representing document, document related information storage unit for storing associated document ID and revision source pointer for each related document, and document name for a plurality of documents A list output unit that outputs a document list including the document list, and the list output unit is revised in one of the documents included in the document list in the document related information storage unit If the pointer is associated with the document list, it outputs the revised notification display indicating the presence of a revised origination pointer for the document.

The document revision management system refers to the revision processing pointer that receives the revision handling operation for the revision notice display and the revision source pointer for the revision notice display according to the revision handling operation, and determines the document name of the corresponding revision target document. A revision source display unit for displaying may be provided.
The revision notification unit may invalidate the revision source pointer related to the revision notification display according to the content of the revision handling operation.
The revision handling operation includes either a revision instruction in which the related document related to the revision notification display is a new revision target document or a revision unnecessary instruction indicating that no revision is made to the related document related to the revision notification display. The unit may invalidate the revision source pointer related to the revision notification display in accordance with either the revision instruction or the revision unnecessary instruction.
The related document specifying unit extracts a candidate document that is a candidate for the related document, and receives a selection instruction regarding whether or not to identify the candidate document as a related document for each of the candidate extracting unit and the candidate document. May be provided.
The candidate extraction unit includes a keyword search unit, the keyword search unit accepts input of a revised keyword according to a revision instruction, and the keyword search unit calculates a revised keyword relevance frequency based on the revised keyword, The extraction unit may extract a candidate document related to the revised keyword based on the revised keyword relevance frequency.
The keyword search unit calculates the revised keyword weight for each revised keyword based on the content of the revised document, and the keyword search unit revises each revised keyword for each document except the revised document. The revision keyword relevance frequency may be calculated based on the keyword weight.
The keyword search unit may calculate a keyword weight at the time of revision based on a character string immediately before and after the keyword at the time of revision appearing in the document to be revised.
The keyword search unit calculates, for each of the documents excluding the document to be revised, the revision keyword hit count indicating the number of times the revision keyword appears for each revision keyword based on the content of the document. The keyword search unit may calculate the revision-related keyword relevance frequency based on the revision keyword hit count for each of the revision keywords for each of the documents excluding the revision target document.
The candidate extraction unit includes a keyword search unit, the keyword search unit accepts input of keywords at the time of registration, the document related information storage unit stores the document in association with the keywords at the time of registration, and the keyword search unit registers The registration keyword relevance frequency may be calculated based on the current keyword, and the candidate extraction unit may extract a candidate document related to the registration keyword based on the registration keyword relevance frequency.
The keyword search unit calculates a registration keyword weight for each of the registration keywords for each document based on the contents of the document, and the document related information storage unit calculates the registration keyword and the registration keyword weight. The keyword search unit may calculate the keyword relevance frequency at the time of registration based on the keyword weight at the time of registration for each of the documents excluding the revision target document.
The keyword search unit may calculate a keyword weight for registration based on the keyword highlighting symbol immediately before and the keyword highlighting symbol immediately after the registration keyword appearing in the document.
The keyword search unit calculates, for each of the documents excluding the document to be revised, the number of registered keyword hits that represents the number of times the registered keyword appears for each registered keyword based on the content of the document. The keyword search unit may calculate the keyword relevance count at the time of registration based on the number of keyword hits at the time of registration for each of the documents excluding the document to be revised.
The document revision management system further accepts input of a creator ID that represents the creator of the document and classification information that represents the classification of the document, and the candidate extraction unit selects the document name, creator, among the documents excluding the document to be revised. A document in which at least one of ID and classification information matches the revision target document may be extracted as a candidate document related to the attribute.
The candidate extraction unit further includes a reference analysis unit for identifying a reference name related to the document to be revised, and the candidate extraction unit is a document excluding the document to be revised, in which the reference name appears in the document. May be extracted as a candidate document related to the reference document name.
The bibliography analysis unit identifies a specific character string indicating the boundary between the text of the document to be revised and the reference name list, and based on the reference highlighting symbol that appears after the specific character string indicating the boundary. A literature name may be specified.
The reference document analysis unit may specify the reference document name based on the reference document highlighting symbol and a specific reference document character string that appears after the reference document highlighting symbol.

  The document revision management program according to the present invention causes a computer to function as the above-described document revision management system.

  According to the document revision management system and the document revision management program according to the present invention, the related document identification unit identifies a related document related to the revision target document, the revision notification unit creates a revision source pointer representing the revision target document, Since the list output unit outputs a revision notification indicating the existence of the revision source pointer to the document list, even a general viewer can check the revision notification display, and the related document needs to reflect the revision. You can know that there is.

It is a figure which shows the structure containing the document revision management system which concerns on Embodiment 1 of this invention. It is a figure which shows the example of a structure of the document memory | storage part of FIG. It is a figure which shows the example of a structure of the document relevant-information memory | storage part of FIG. It is a figure which shows the example of a structure of the highlight symbol setting file of FIG. It is a figure which shows the example of a structure of the reference literature determination setting file of FIG. 3 is a flowchart showing a processing flow when the document revision management system of FIG. 1 registers (stores) a new document in a document storage unit. It is a figure which shows the example of the document registered into the document memory | storage part of FIG. It is a flowchart which shows the detail of step S12 of FIG. It is a figure which shows the example of the attribute search screen displayed by FIG.6 S12. It is a flowchart which shows the detail of step S13 of FIG. It is a figure which shows the example of the keyword search screen displayed by FIG.6 S13. It is a flowchart which shows the detail of step S14 of FIG. It is a figure which shows the example of the selection instruction | indication input screen displayed by FIG.6 S14. 3 is a flowchart showing a flow of processing when the document revision management system of FIG. 1 accepts a document revision operation. It is a flowchart which shows the detail of step S22 of FIG. It is a flowchart which shows the detail of step S23 of FIG. It is a flowchart which shows the detail of step S24 of FIG. 2 is a flowchart showing a flow of processing when the document revision management system of FIG. 1 displays a revision notification in association with document list display. It is a figure which shows the example of a structure of the document list screen displayed by FIG.18 S31.

Embodiments of the present invention will be described below with reference to the accompanying drawings.
Embodiment 1 FIG.
FIG. 1 shows a configuration including a document revision management system 100 according to Embodiment 1 of the present invention. The document revision management system 100 can be configured using, for example, an integrated document management system “Manage Leader (registered trademark)” (manage reader) provided by Mitsubishi Electric Information Systems Corporation.
The document revision management system 100 has a configuration as a well-known computer, and includes a calculation unit 10 that performs calculation and a storage unit 30 that stores information. The computing means 10 includes a CPU (Central Processing Unit), and the storage means 30 includes a storage medium such as a semiconductor memory and an HDD (Hard Disk Drive).

  Although not particularly illustrated, the document revision management system 100 includes an input unit, an output unit, and a network interface. The input means is means used by the user to input information, and is, for example, a mouse or a keyboard. The output means is means for outputting information to the user, and is a display device such as a liquid crystal display, but may be a printing device such as a printer. The network interface is a means for inputting / outputting information to / from the network to which the document revision management system 100 is connected. The document revision management system 100 is connected to the communication network 40 through the network interface so as to be communicable.

  The communication network 40 is a LAN, for example, but may be a WAN such as the Internet. A plurality of user terminals 50 are connected to the communication network 40. Similar to the document revision management system 100, the user terminal 50 has a configuration as a well-known computer, and includes a calculation unit, a storage unit, an input unit, an output unit, and a network interface. With such a configuration, the document revision management system 100 and the user terminal 50 can communicate with each other via the communication network 40.

  The storage means 30 of the document revision management system 100 includes a document storage unit 31 and a document-related information storage unit 32, and stores a highlight symbol setting file 33 and a reference determination setting file 34.

  The document storage unit 31 stores a plurality of documents. The document storage unit 31 can be designed as a document database having a known configuration. Further, the document storage unit 31 may be configured as a document management system having a function as an independent computer by itself, and in that case, the document storage unit 31 can communicate with other computers included in the document revision management system 100. Also good.

  FIG. 2 shows an example of the configuration of the document storage unit 31. The document storage unit 31 stores, for a plurality of documents, a document ID that uniquely identifies each document and information related to the document in association with each other. FIG. 2 shows an example of five documents, and information relating to the document includes “document name”, “attribute”, and “content”. For example, the document name of the document whose document ID is 001 is “A document”. The “attribute” includes a registration date, a creator ID, and classification information. The registration date represents the date when the document is stored in the document storage unit 31, and the creator ID uniquely identifies the creator of the document (or the registrant who stored the document in the document storage unit 31). It represents a user ID, and the classification information represents information such as a character string or a symbol string arbitrarily designated by the creator to classify the document.

  FIG. 3 shows an example of the configuration of the document related information storage unit 32. For each of the documents stored in the document storage unit 31, the document related information storage unit 32 stores the document ID and information related to the relationship between the document and other documents in association with each other. Information relating to relevance includes keyword information, a related document pointer, and a revision source pointer.

  The keyword information is information relating to a keyword (keyword at the time of registration) specified when the document is registered in the document storage unit 31 as a keyword for evaluating the relevance between the document and another document. Specifically, the keyword information includes “registration keyword” and “registration keyword weight” calculated for each of the registration keywords, which are associated with each other. In the example of FIG. 3, for the document with the document ID “002”, the registration keyword “Einstein” is associated with a value “1.3” as the registration keyword weight, and the registration keyword “physics” A value of “1.4” is associated as a keyword weight at the time of registration.

  The related document pointer is a pointer indicating that the document is related to another document. In the example of FIG. 3, the document with the document ID “001” is related to the document with the document ID “004”. This association is not necessarily bidirectional. For example, in the example of FIG. 3, the document with the document ID “004” is not related to the document with the document ID “001”.

  The revision source pointer is a pointer that indicates the revision source document when there is a possibility that revision of another document may be necessary due to revision of a document (the revision source document) For example, it is represented by the document ID of the document to be revised. In the example of FIG. 3, the revision source pointer “001” is associated with the document with the document ID “003”. This is because the document with the document ID “001” has been revised. ”Indicates that there is a possibility that revision will be necessary.

  FIG. 4 shows an example of the configuration of the highlight symbol setting file 33. The highlight symbol setting file 33 is a file that defines a highlight symbol for identifying a character string highlighted in a document and a highlight symbol weight that represents the degree of emphasis of the character string. The highlight symbol setting file 33 associates highlight symbol weights with respect to one or more types of highlight symbols. The highlighting symbol includes an emphasis start character that indicates the start of the character string and an emphasis end character that indicates the end of the character string. In the example of FIG. 4, the highlight symbols include key brackets, double key brackets, cornered key brackets, double quotation marks, and rhombus symbols.

In the example of FIG. 4, if there is a pattern in the document that “closed double brackets appear after the open double brackets”, the character string between the open double brackets and the closed double brackets A highlight symbol weight of 1.2 is given.
As will be described later, in the present embodiment, the highlighting symbol in FIG. 4 is also treated as a keyword highlighting symbol corresponding to the keyword highlighting, and the reference highlighting corresponding to the highlighting of the reference document. It is also treated as a display symbol. However, as a modification, the keyword highlighting symbol and the reference highlighting symbol may be defined independently.

  FIG. 5 shows an example of the configuration of the reference document determination setting file 34. The reference determination setting file 34 is a file that defines a reference reference character string that may represent a reference to a reference. The reference text string includes “according to”, “according to”, “seeing” and “in”. For example, if a character string “According to” appears in a document, the character string immediately before that may represent a reference to a reference document. As will be described later, in the present embodiment, only when a character string corresponding to the highlighted symbol in FIG. 4 appears immediately before the reference character string in FIG. 5, the character string is recognized as a reference document name.

Although not shown, the storage unit 30 stores a document revision management program. This document revision management program causes a computer to function as the document revision management system 100.
The computing means 10 of the document revision management system 100 functions as a document relation management unit that manages the relation of documents by executing this document revision management program. More specifically, the computing unit 10 functions as a document input unit 11, a related document specifying unit 12, a list output unit 13, a revision instruction receiving unit 14, a revision notification unit 15, a pointer process receiving unit 16, and a revision source display unit 17. To do. Further, the function as the related document specifying unit 12 includes a function as the candidate extracting unit 20 and a function as the selection instruction receiving unit 25. Further, the function as the candidate extraction unit 20 includes a function as the attribute search unit 21, a function as the keyword search unit 22, and a function as the reference document analysis unit 23.
Details of these functions will be described below.

<I. When registering a new document>
FIG. 6 is a flowchart showing the flow of processing when the document revision management system 100 registers (stores) a new document in the document storage unit 31.
First, the document input unit 11 accepts input of a document (step S11). Specifically, the document input unit 11 accepts input of document contents and document names. Here, it is assumed that the document shown in FIG. 7 is input with the document name “X document”. In the X document, three documents, α document, β document, and γ document, are cited as reference documents, and these correspond to the reference document names appearing in the A document, C document, and D document in FIG. 2, respectively. ing. There is also a reference to the δ document in the content of the document.
In step S11, the document input unit 11 generates a document ID for uniquely identifying a new document in response to the input of the document.

  Here, the input of the document and the document name is performed from the user terminal 50 via the communication network 40, for example. Although not specifically described below, other information can be input to the document revision management system 100 in the same manner. The input of information in the user terminal 50 can be performed by executing a known web browser program, for example.

Next, the related document specifying unit 12 specifies a related document based on the attribute (step S12).
FIG. 8 shows details of step S12. Step S12 includes steps S121 to S125 as more detailed steps.
In step S12, the attribute search unit 21 first displays an attribute search screen (step S121). FIG. 9 shows an example of this attribute search screen. However, the example of FIG. 9 is the one at the time when the processing up to step S124 is completed.

  Here, the display of the attribute search screen is performed on the user terminal 50 via the communication network 40, for example. For example, the document revision management system 100 creates data representing an attribute search screen (HTML format file or the like) and transmits it to the user terminal 50. The user terminal 50 receives this and displays the corresponding screen. . Although not specifically described below, display of other screens and output of other information performed by the document revision management system 100 can be performed in the same manner. The display of information on the user terminal 50 can be performed, for example, by executing a known web browser program.

  Next, the attribute search unit 21 receives input of attributes of a new document (X document in this example) (step S122). In the example of FIG. 9, the symbol “ABC1” is input as the classification information. Further, the attribute search unit 21 determines a registration date and a creator ID as attributes of the X document. The registration date is the date when step S11 was executed, and the creator ID is the user ID of the user who entered the X document. Further, the document name input in step S11 is also handled as the attribute of the document.

  Next, the attribute search unit 21 searches for a document in the document storage unit 31 based on the input attribute, and extracts a document that matches any of the attributes as a candidate document (step S123). Here, the “candidate document” is a document stored in the document storage unit 31 as a candidate that may be related to the document newly input in step S11, that is, a candidate that may be a related document. Means a document to be selected and presented to the creator of the new document.

  In this example, a document whose registration date is January 1, 2011 is searched as an attribute, a document whose creator ID is “U09” is searched, and a document whose classification information is “ABC1” is searched. Will do. Among the documents in FIG. 2, the classification information of the A document, the C document, and the E document is “ABC1”, which corresponds to this, and is extracted as a candidate document. (In this example, it is assumed that there is no candidate document having the same registration date and creator ID.)

  Next, the selection instruction receiving unit 25 displays a list of candidate documents and receives a selection instruction regarding whether or not each candidate document is to be identified as a related document (step S124). In the example of FIG. 9, for each candidate document, a check box is displayed in the “Related” column, and the creator can arbitrarily check these check boxes. The operation of checking this check box corresponds to a selection instruction for specifying the candidate document as a related document. Similarly, an operation of not checking this check box (or an operation of removing the check) corresponds to a selection instruction not to make the candidate document a related document. Here, only the C document is selected as the related document.

  Next, the selection instruction receiving unit 25 creates a related document pointer based on the selection instruction and stores it in the document related information storage unit 32 (step S125). For example, in the example of the selection instruction shown in FIG. 9, “003”, which is the document ID of the C document, is stored in association with the X document as the related document pointer.

Note that the processing in steps S122 to S125 in FIG. 8 is executed in parallel or sequentially by the number of attributes. That is, the same processing is executed for the creator ID and the registration date, and a document by the same creator and a document registered on the same date are displayed as candidate documents. In addition, when additional attributes are defined in addition to these, the same processing is executed for these attributes.
The above is the detail of step S12.

  After step S12, the related document specifying unit 12 specifies the related document based on the registration keyword (step S13).

FIG. 10 shows details of step S13. Step S13 includes steps S131 to S139 as more detailed steps.
In step S13, the keyword search unit 22 first displays a keyword search screen (step S131). FIG. 11 shows an example of this keyword search screen. In addition, the example of FIG. 11 is a thing at the time of the process to step S138 being complete | finished.

Next, the keyword search unit 22 receives an input of a keyword when registering a new document (X document in this example) (step S132). In the example of FIG. 11, a character string “physical” is input as a registration keyword.
Next, the keyword search unit 22 performs a full-text search of the X document based on the input registration keyword, and specifies all appearance positions (step S133).

Next, the keyword search unit 22 calculates a registration keyword weight for each registration keyword (step S134). If the registration keyword weight is W1, W1 is calculated according to the following equation.
W1 = 0.1 × [number of hits] + [sum of highlighted symbol weights]
Here, the number of hits is the number of hits of the registered keyword in the X document, and the sum of the highlight symbol weights is the highlight symbol weight corresponding to the highlight symbol (see FIG. 4) at each appearance position. Is the sum of the keywords at the time of registration. For example, if the character string “physical” appears 10 times in an X document, 2 of which have double quotation marks and 3 have double key brackets, the number of hits is 10, and the highlight symbol The sum of the weights is (2 × 0.2) + (3 × 1.2) = 4.0. Therefore, in this case, the registration keyword weight associated with the registration keyword “physical” is W1 = 0.1 × 10 + 4.0 = 5.0. In this way, the keyword search unit 22 calculates the registration keyword weight for each of the registration keywords based on the contents of the newly registered X document. In this way, the keyword search unit 22 registers a newly registered X document based on the keyword highlighting symbol immediately before and the keyword highlighting symbol immediately after the registration keyword appearing in the document. Hour keyword weight is calculated.
Further, the keyword search unit 22 stores the calculated registration keyword weight in the document related information storage unit 32.

  Next, the keyword search unit 22 performs full-text search for other documents (all documents stored in the document storage unit 31) based on the registered keyword, and measures the hit number H1 for each document (step S135). .

Next, the keyword search unit 22 calculates the degree of association with the document (X document) for all the documents stored in the document storage unit 31 (step S136). The relevance frequency calculated here is assumed to be the initial relevance frequency R1. R1 is calculated according to the following equation for each document.
R1 = Σ _keyword (W1 · H1)
Here, as described above, W1 is a registration keyword weight for one registration keyword, and H1 is the number of times the registration keyword is hit in the document. When this product is summed up for all the keywords at the time of registration, an initial association degree R1 between the X document and the document is obtained.

  Next, the keyword search unit 22 extracts candidate documents based on the initial relevance frequency R1 of each document (step S137). For example, a predetermined number of documents may be set as candidate documents in descending order of the value of R1, or documents included in a predetermined ratio from the higher value of R1 may be set as candidate documents.

Next, the selection instruction receiving unit 25 displays a list of candidate documents as in step S124 of FIG. 8 and issues a selection instruction regarding whether or not each candidate document is specified as a related document for each candidate document. Accept (step S138).
Next, the selection instruction receiving unit 25 creates a related document pointer based on the selection instruction and stores it in the document related information storage unit 32, similarly to step S125 of FIG. 8 (step S139).

As described above, steps S135 and S136 of FIG. 10 are executed in parallel or sequentially for each of the documents stored in the document storage unit 31. Also, steps S132 to S134 in FIG. 10 are executed in parallel or sequentially for each of the input registration keywords.
The above is the details of step S13.

  After step S13, the related document specifying unit 12 specifies a related document based on the reference document name (step S14).

FIG. 12 shows details of step S14. Step S14 includes steps S141 to S144 as more detailed steps.
In step S14, the reference document analysis unit 23 first identifies a reference document name related to the X document (step S141). In step S141, the reference document analysis unit 23 specifies a reference document name using the following two criteria as a reference document name determination criterion.

  The first criterion assumes a sentence structure in which a reference name list is collected at the end of a new document (X document in this example), and specifies a reference name by paying attention to the reference name list. It is a standard for. According to the first standard, the reference document analysis unit 23 uses a specific character string indicating the boundary between the text and the reference document name list, and the highlighted symbol that appears after the character string indicating the boundary (FIG. 4). It is determined that the character string enclosed in (see) is a reference name.

For example, as the specific character string indicating the boundary between the text and the reference document name list, the last character string of “reference documents” appearing in the document can be used. That is, the reference document analysis unit 23 stores the last hit part of the character string “reference document” appearing in the X document, and after that part, the character string surrounded by the highlighted symbols (that is, It is determined whether or not an emphasis start character, some character string, and an emphasis end character sequence) appear. Then, if there is such a character string, the corresponding character string is specified as a reference document name.
For example, in the X document of FIG. 7, after the character string “reference document” that appears last, an open double parenthesis, a character string “α literature”, and a closed double key bracket are in this order. The character string “α document” is specified as the reference document name. The same applies to “β literature” and “γ literature”.

  The second standard is a standard for identifying a reference name based on a highlight symbol and a specific character string indicating that a reference name is described (reference reference character string, see FIG. 5). It is. That is, the reference document analysis unit 23 determines whether or not a sequence in the order of an emphasis start character, some character string, an emphasis end character, and a reference reference character string appears in the X document. If there is, the corresponding character string is identified as the reference name. For example, in the X document of FIG. 7, an open key bracket, a character string “δ literature”, a closing key bracket, and a character string “in” appear in this order as a string. The column is identified as the reference name.

  Next, the reference document analysis unit 23 performs a full text search for all the documents stored in the document storage unit 31 for each identified reference document name, and extracts candidate documents (step S142). For example, it is determined whether or not a character string “α document” appears in the A document to the E document, and if there is an occurrence, the document is set as a candidate document. In this example, the reference name “α document” appears in the A document, the reference name “β document” appears in the C document, and the reference name “γ document” appears in the D document. Since they appear, the A document, the C document, and the D document are extracted as candidate documents.

  Next, the selection instruction receiving unit 25 displays a list of candidate documents as in step S124 of FIG. 8 and issues a selection instruction regarding whether or not each candidate document is specified as a related document for each candidate document. Accept (step S143). FIG. 13 shows an example of this selection instruction input screen. In the example of FIG. 13, the list of candidate documents is classified and displayed for each reference document name. In this state, “β document” is designated as the reference document name, and a list of candidate documents including the reference document name ( In this case, only the C document) is displayed.

Next, the selection instruction receiving unit 25 creates a related document pointer based on the selection instruction and stores it in the document related information storage unit 32, similarly to step S125 of FIG. 8 (step S144).
As described above, step S142 in FIG. 12 is executed in parallel or sequentially for each reference name.
The above is the detail of step S14.

  After step S14, the document input unit 11 stores the document ID, document name, and contents in the document storage unit 31 (step S15). If this process is necessary for the creation of the related document pointer, the process of step S15 or a part thereof may be appropriately executed before the creation of the related document pointer.

<II. Document revision>
FIG. 14 is a flowchart showing a processing flow when the document revision management system 100 accepts a document revision operation.
First, the revision instruction receiving unit 14 receives a revision instruction for any of the documents stored in the document storage unit 31 (step S21). Hereinafter, the document subject to the revision instruction is referred to as “revision target document”. That is, a user (revision person) who intends to revise a document selects a revision target document by designating a document name, a document ID, etc., for example, and instructs the document revision management system 100 to start the processing in FIG. .
In the following description, as an example, it is assumed that an X document having the contents shown in FIG. 7 is stored in the document storage unit 31 and this X document is selected as a revision target document.

  Next, the related document specifying unit 12 specifies a related document related to the revision instruction based on the attribute (step S22). As can be seen from the following description, the related document specified in step S22 may be the same as the related document specified at the time of document registration (see step S12 in FIG. 6), and a part thereof overlaps. Sometimes it may be completely different.

FIG. 15 shows details of step S22. Step S22 includes steps S221 to S225 as more detailed steps.
In step S22, the attribute search unit 21 first displays an attribute search screen (step S221). The attribute search screen is the same as the attribute search screen at the time of document registration shown in FIG. 9, for example.

Next, the attribute search unit 21 receives input of attributes of the revision target document (X document in this example) (step S222). Here, the attribute search unit 21 may use the attribute designated at the time of registration of the X document as it is in accordance with the operation of the revision person. In addition, the attributes of the already specified X document may be changed according to the operation of the revision person, or attributes may be added.
Here, it is assumed that the attribute of the X document is not changed and the attribute of FIG. 9 is used as it is.

  Next, the attribute search unit 21 searches for documents in the document storage unit 31 based on the attributes of the revision target document and extracts candidate documents, similarly to step S123 of FIG. 8 (step S223). That is, the attribute search unit 21 extracts a document excluding the revision target document that has at least one of the document name, creator ID, and classification information that matches the revision target document as a candidate document related to the attribute.

  Next, the selection instruction receiving unit 25 displays a list of candidate documents as in step S124 of FIG. 8 and issues a selection instruction regarding whether or not each candidate document is specified as a related document for each candidate document. Accept (step S224). By this selection instruction, the revision person can arbitrarily designate a related document.

Next, based on the selection instruction, the revision notification unit 15 creates a revision source pointer and stores it in the document related information storage unit 32 (step S225). The revision source pointer includes information indicating the revision target document. For example, in the example of the selection instruction shown in FIG. 9, for the C document, the document ID of the X document is associated and stored as the revision source pointer. (Note that the related document pointer created at the time of new registration of the document includes the document ID of the related document and is associated with the newly registered document, but the revision source pointer created at the time of revision is a reverse pointer, (The document ID of the document to be revised is included and associated with the related document.)
The above is the detail of step S22.

After step S22, the related document specifying unit 12 specifies a related document related to the revision instruction based on the keyword (step S23).
FIG. 16 shows details of step S23. Step S23 includes steps S2301 to S2315 as more detailed steps. Of these, steps S2302 to S2309 represent processing for specifying a related document based on a revision keyword specified at the time of revision. Steps S2310 to S2315 represent processing for specifying a related document based on a registration keyword specified at the time of registration and stored in the document related information storage unit 32.

  In step S23, the keyword search unit 22 first displays a keyword search screen (step S2301). The keyword search screen is the same as the keyword search screen at the time of document registration shown in FIG. 11, for example.

Next, the keyword search unit 22 accepts an input of a keyword at the time of revision according to the revision instruction (step S2302). The revision keyword is arbitrarily determined by the revision person in consideration of the revision content, for example.
Next, the keyword search unit 22 performs a full-text search for the document to be revised based on the inputted revision keyword, and specifies all appearance positions (step S2303).

Next, the keyword search unit 22 calculates a revised keyword weight for each revised keyword (step S2304). When the revised keyword weight is W2, W2 is calculated according to the following equation.
W2 = 0.1 × [number of hits] + [sum of highlighted symbol weights]
This calculation method is the same as the calculation method of the registration keyword weight W1 for the registration keyword. In this way, the keyword search unit 22 calculates the revision keyword weight for each revision keyword based on the contents of the revision target document. Further, the keyword search unit 22 calculates a revision keyword weight for the revision target document based on the keyword highlighting symbol immediately before the revision keyword appearing in the document and the keyword highlighting symbol immediately after the keyword.

  Next, the keyword search unit 22 performs full-text search on other documents (all documents except the document to be revised among documents stored in the document storage unit 31) based on the revision keyword, and hits each document. The number H2 is measured (step S2305).

Next, the keyword search unit 22 calculates the degree of association with the revision target document for each of the other documents (step S2306). The relevance frequency calculated here is the keyword relevance frequency R2 at the time of revision. R2 is calculated according to the following equation for each document.
R2 = Σ _keyword (W2 · H2)
Here, as described above, W2 is the revision keyword weight for one revision keyword, and H2 is the number of times the revision keyword is hit in the document. When this product is summed up for all revision keywords, the revision keyword relevance frequency R2 between a certain document and the revision target document is obtained.

  Next, the keyword search unit 22 extracts candidate documents based on the revision-related keyword association frequency R2 of each document (step S2307). For example, a predetermined number of documents may be set as candidate documents in descending order of the value of R2, or documents included in a predetermined ratio from those having a larger value of R2 may be set as candidate documents.

Next, the selection instruction accepting unit 25 displays a list of candidate documents as well as the step S224 of FIG. 15 and issues a selection instruction regarding whether or not each candidate document is to be specified as a related document. Accept (step S2308).
Next, the revision notification unit 15 creates a revision source pointer based on the selection instruction and stores the revision source pointer in the document related information storage unit 32 as in step S225 of FIG. 15 (step S2309).

Next, the keyword search unit 22 refers to the document related information storage unit 32, and acquires all the registration keywords and registration keyword weights of the document to be revised (step S2310).
Next, based on the acquired registration keyword, the keyword search unit 22 performs a full-text search for other documents (all documents other than the document to be revised among the documents stored in the document storage unit 31). The number of hits H3 is measured (step S2311).

Next, the keyword search unit 22 calculates the degree of association with the revision target document for each of the other documents (step S2312). The relevance frequency calculated here is set as a keyword relevance frequency R3 at the time of registration. R3 is calculated according to the following equation for each document.
R3 = Σ _keyword (W1 · H3)
Here, W1 is a registration keyword weight for one registration keyword, and is a value calculated and recorded at the time of document registration. H3 is the number of times that the keyword at the time of registration is hit in the document, and is the value measured in step S2311. When this product is summed up for all registration keywords, the registration-related keyword relevance frequency R3 between a certain document and the revision target document is obtained.

  Next, the keyword search unit 22 extracts candidate documents based on the keyword-related frequency R3 at the time of registration of each document (step S2313). For example, a predetermined number of documents may be set as candidate documents in descending order of the value of R3, or documents included in a predetermined ratio may be set as a candidate document in descending order of the value of R3.

Next, the selection instruction accepting unit 25 displays a list of candidate documents as well as the step S224 of FIG. 15 and issues a selection instruction regarding whether or not each candidate document is to be specified as a related document. Accept (step S2314).
Next, the revision notifying unit 15 creates a revision source pointer based on the selection instruction and stores it in the document related information storage unit 32, similarly to step S225 of FIG. 15 (step S2315).
The above is the details of step S23.

  In this way, in step S23, candidate documents are extracted using not only the registration keyword but also the revision keyword input in connection with the revision instruction. Therefore, the registration keyword and initial keyword specified at the initial registration are extracted. Regardless of the related document selected at the time of registration, a new related document can be selected for each revision according to the content of the revision.

After step S23, the related document specifying unit 12 specifies a related document related to the revision instruction based on the reference document name (step S24).
FIG. 17 shows details of step S24. Step S24 includes steps S241 to S244 as more detailed steps.

  In step S24, the reference document analysis unit 23 first specifies the name of the reference document related to the document to be revised based on the first standard and the second standard in the same manner as in step S141 in FIG. 12 (step S241). . Using the first standard, the reference document analysis unit 23 displays a highlighted symbol (see FIG. 4) that appears after the character string “reference document” that is a specific character string indicating the boundary between the text and the reference name list. The character string enclosed in (see) is specified as the bibliography name. Also, using the second standard, the reference name is identified based on the highlighted symbol and the specific character string indicating that the reference name is described (reference reference character string, see FIG. 5). To do.

  Next, the reference document analysis unit 23, for each of the identified reference document names, converts all the documents (except the document to be revised) stored in the document storage unit 31 in the same manner as in step S142 of FIG. Search and extract candidate documents (step S242).

Next, the selection instruction accepting unit 25 displays a list of candidate documents as well as the step S224 of FIG. 15 and issues a selection instruction regarding whether or not each candidate document is to be specified as a related document. Accept (step S243).
Next, the revision notification unit 15 creates a revision source pointer based on the selection instruction and stores it in the document related information storage unit 32 as in step S225 of FIG. 15 (step S244).
The above is the details of step S24.

Here, since the extraction of candidate documents in steps S22 to S24 covers all documents stored in the document storage unit 31 when the revision instruction is input, the document storage unit 31 is later than the revision target document. Registered documents can also be considered.
In addition, since related documents can be specified again at the time of revision, regardless of the related documents specified at the time of registration, if the important candidate documents differ depending on the content of the revision, an appropriate related document depending on the situation Can be specified.

  After step S24, the document input unit 11 executes a revision process for the revision target document (step S25). This revision process may be an update process using a function of a well-known document editing software, for example. When the revision process is completed, the revised document to be revised is stored in the document storage unit 31. By this revision process, the revision person can actually revise the content of the document to be revised.

  Note that the above is processing when a revision instruction is input for one revision target document, but the revision instruction can be input independently for different documents. In addition, one document may be specified as a related document related to multiple revision target documents. In this case, multiple revision source pointers are associated with the document, and each revision source pointer has a different revision target document. Will represent.

<III. When revision notice is displayed>
FIG. 18 is a flowchart showing the flow of processing when the document revision management system 100 displays a revision notification in relation to the document list display.
First, the list output unit 13 outputs a document list of documents stored in the document storage unit 31 (step S31). FIG. 19 shows an example of the configuration of a document list screen that displays this document list. The output document list is displayed on the user terminal 50, for example.

The document list includes a document name of each document and a revision notification display indicating that a revision source pointer is associated with each document. In the example of FIG. 19, the “notification” column and the “document name” column are included in one row, and the “notification” column includes a check box. A check mark is an example of a revision notification display. When a check mark is displayed in a check box, it indicates that a revision source pointer is associated with a document represented by a document name displayed on the same line. In this example, revision source pointers are associated with the B document, the C document, and the D document, respectively.
As described above, when the revision source pointer is associated with any of the documents included in the document list in the document related information storage unit 32, the list output unit 13 indicates the existence of the revision source pointer for the document in the document list. A revision notification display is output.

  Here, the “document list” does not necessarily have to be a list of all documents stored in the document storage unit 31, but may be a list of only documents that meet a specific condition. For example, it may be a list of documents classified in a specific folder in the document storage unit 31, a list of documents belonging to a specific field, a list of documents including a specific keyword, and the like.

  As users who may become viewers of the document list, all users who want to refer to any of the documents stored in the document storage unit 31 are applicable. That is, it includes a user who becomes the creator of any document and a user who is not the creator of any document. Therefore, even a general reader who is not the creator of any document can check the revision notice display by browsing the document list, and the related document needs to reflect the revision (or At least, the necessity of reflecting the revision has not been determined). For this reason, the viewer does not misunderstand that the state before reflection is the latest content.

After step S31, the pointer process accepting unit 16 accepts an operation for the revision notification display included in the document list (step S32). This operation is, for example, a click operation on the check mark in FIG. 19 and constitutes a part of the revision handling operation.
Next, the revision source display unit 17 acquires the revision source pointer related to the operated revision notification display in accordance with the revision handling operation (step S33). For example, when a check mark corresponding to the B document in FIG. 19 is clicked, the revision source pointer associated with the B document is acquired. Hereinafter, the document (in this example, the B document) associated with the operated revision notification display is referred to as a “revision necessity determination document”. Note that this revision necessity determination document is designated as a related document when viewed from the revision target document indicated by the revision source pointer.

Next, the revision source display unit 17 displays the document name of the document (revision target document) corresponding to the acquired revision source pointer (step S34). Specifically, the revision source display unit 17 refers to the document related information storage unit 32 using the document ID included in the revision source pointer of the document for which revision necessity determination is being performed, and determines the document name associated with the document ID. Get and display this.
Here, although not shown in FIG. 18, the viewer of the document list refers to the revision content of the revision target document, and determines whether the revision of the document requiring revision is necessary according to the revision content. .

Thereafter, the pointer process accepting unit 16 accepts an instruction as to whether or not to revise the document for which revision is necessary (B document in this example) (step S35). This instruction forms part of the revision handling operation. The instruction indicating that the revision is to be performed is, for example, a separate revision instruction in which the document requiring revision is a new revision target document. In this case, in response to the revision instruction, the process of FIG. 14 is recursively executed for the document (the entire process of FIG. 14 is represented as step S36).
Although not shown in FIG. 18, the instruction accepted in step S35 is only an instruction from a specific user. For example, only when the authority to revise a document (revise authority) is given to one or more user IDs, and the viewer's user ID has the authority to revise, the pointer processing reception unit 16 instructs whether or not to make a revision. Accept. On the other hand, if the user ID of the viewer does not have revision authority, the pointer process accepting unit 16 ignores even if an instruction to perform revision is input (that is, does not execute steps S35 to S37). . This revision authority may be specified individually for each document. In addition, when the viewer of the document that is being judged as being revised is the creator of the document, the pointer processing reception unit 16 may treat the viewer as having the authority to revise the document.

After step S36 is executed, or when an instruction indicating that no revision is to be made (revision unnecessary instruction) is input in step S35, the revision notification unit 15 invalidates the revision source pointer (step S37). . That is, the revision notification unit 15 invalidates the revision source pointer related to the revision notification display that is the target of the revision handling operation in accordance with either the revision instruction or the revision unnecessary instruction for the document requiring revision.
Here, “invalidating the revision source pointer” is performed, for example, by deleting the revision source pointer from the document related information storage unit 32. Alternatively, information indicating that the revision source pointer is invalid (such as a flag) may be associated with the revision source pointer.

  As described above, according to the document revision management system 100 according to the first embodiment, since the contents of revision of a document are not reflected in other related documents, even a general viewer can know. The revision notification display can be confirmed by browsing the document list, and it is possible to know that the related document needs to reflect the revision.

  In addition, the creator of the related document can find the revision notice display by browsing the document list and notice the possibility that the document needs to be revised. it can. In addition, when extracting candidate documents, by using a full-text search using keywords, it is possible to eliminate omissions in association and reliably reflect the revised contents.

The following modifications can be made to the document revision management system 100 according to the first embodiment.
The internal configuration of the calculation means 10 may be different from that shown in FIG. 1, and the function sharing of the components of the calculation means 10 may be different from that of the first embodiment. For example, the selection instruction receiving unit 25 may not be provided. In that case, for example, the attribute search unit 21 may execute step S124 of FIG.

  The structural unit of hardware is not limited to that shown in the first embodiment. For example, the document revision management system 100 may be configured by a plurality of computers. For example, the document storage unit 31 may be configured as an independent computer. Alternatively, the document revision management system 100 and the user terminal 50 may be configured as a single computer.

  The creation timing of the related document pointer and the creation timing of the revision source pointer may be different from those in the first embodiment. For example, step S225 in FIG. 15, steps S2309 and S2315 in FIG. 16, and step S244 in FIG. 17 may be combined into one step and executed after step S24 in FIG.

  In Embodiment 1, in the process at the time of revision, the keyword search unit 22 refers to the keyword at the time of registration, but the candidate extraction unit 20 may use the information at the time of registration other than this. For example, the attribute search unit 21 may refer to the attribute stored in the storage unit 30 without accepting the input of the revision person in step S222 of FIG. The reference document analysis unit 23 stores the reference document name specified at the time of registration (step S141 in FIG. 12), and does not specify the reference document name again at the time of revision (step S241 in FIG. 17). The stored reference name may be used as it is.

  In the first embodiment, in the search for candidate documents based on the attributes (step S123 in FIG. 8 and step S223 in FIG. 15), the search is performed based only on the attributes of the new document or the revision target document. As a modification, in addition to or regardless of the attribute of the new document or the document to be revised, input of an attribute to be used for the search may be accepted. For example, another creator ID may be specified, and a specific period may be specified as a range of registration dates.

  In the first embodiment, the registration-time keyword weight W1 and the revision-time keyword weight W2 are both calculated using the formula of “0.1 × [number of hits] + [sum of highlighted symbol weights]”. This expression may be different, or the expression may be different for the registration keyword weight W1 and the revision keyword weight W2. This expression represents a relative ratio of how much importance is given to [number of hits] and [sum of highlighted symbol weights], and [number of hits] is set to 0.1 in the first embodiment. On the other hand, the [total sum of highlighted symbol weights] is 1. By changing the constant of 0.1, it is possible to arbitrarily adjust the relative importance of [number of hits] with respect to [total weight of highlighted symbol weight].

DESCRIPTION OF SYMBOLS 10 Calculation means (11 Document input part, 12 Related document specific | specification part, 13 List output part, 14 Revision instruction | indication reception part, 15 Revision notification part, 16 Pointer process reception part, 17 Revision origin display part, 20 Candidate extraction part, 21 Attributes Search unit, 22 keyword search unit, 23 reference analysis unit, 25 selection instruction reception unit),
30 storage means (31 document storage unit, 32 document related information storage unit, 33 highlighted symbol setting file, 34 reference determination setting file),
40 communication network, 50 user terminals, 100 document revision management system.

Claims (18)

A document storage unit for storing a document ID and a document name in association with each other for a plurality of documents;
A revision instruction receiving unit that receives a revision instruction for a revision target document selected from the documents;
A related document specifying unit for specifying, as a related document, a document related to the revision instruction from among the documents excluding the document to be revised;
A revision notification unit that creates a revision source pointer including information representing the revision target document for each of the related documents;
A document-related information storage unit that stores the document ID and the revision source pointer in association with each other, for each of the related documents;
A list output unit for outputting a document list including document names for the plurality of documents;
With
When the revision source pointer is associated with any of the documents included in the document list in the document related information storage unit, the list output unit indicates the presence of the revision source pointer for the document in the document list. Document revision management system that outputs a revision notice display. The document revision management system includes:
A pointer process accepting unit that accepts a revision handling operation for the revision notification display;
The revision source display unit that displays the document name of the corresponding revision target document with reference to the revision source pointer related to the revision notification display according to the revision handling operation. Document revision management system.   The document revision management system according to claim 2, wherein the revision notifying unit invalidates the revision source pointer related to the revision notice display according to the content of the revision handling operation. The revision handling operation is:
Including either the revision instruction in which the related document related to the revision notification display is a new revision target document, or the revision unnecessary instruction indicating that the related document related to the revision notification display is not revised,
The document revision management system according to claim 3, wherein the revision notification unit invalidates the revision source pointer related to the revision notification display in accordance with either the revision instruction or the revision unnecessary instruction. The related document specifying unit includes:
A candidate extracting unit that extracts candidate documents that are candidates for the related document;
The document revision management according to any one of claims 1 to 4, further comprising: a selection instruction receiving unit that receives a selection instruction regarding whether or not each candidate document is specified as the related document. system. The candidate extraction unit includes a keyword search unit,
The keyword search unit accepts an input of a keyword at the time of revision according to the revision instruction,
The keyword search unit calculates a revised keyword-related frequency based on the revised keyword,
The document revision management system according to claim 5, wherein the candidate extraction unit extracts the candidate document related to the revision keyword based on the revision keyword association frequency. The keyword search unit calculates a revision keyword weight for each of the revision keywords based on the content of the revision target document,
The keyword search unit according to claim 6, wherein, for each of the documents excluding the document to be revised, the revision keyword relevance frequency is calculated based on the revision keyword weight for each of the revision keywords. Document revision management system.   The document revision management system according to claim 7, wherein the keyword search unit calculates the revision keyword weight based on a character string immediately before and after the revision keyword that appears in the revision target document. The keyword search unit, for each of the documents excluding the document to be revised, represents a revision keyword hit indicating the number of times the revision keyword appears for each revision keyword based on the content of the document. Calculate the number,
The keyword search unit calculates the revision-related keyword relevance frequency based on the revision keyword hit count for each of the revision keywords for each of the documents excluding the revision target document. The document revision management system as described in any one of. The candidate extraction unit includes a keyword search unit,
The keyword search unit accepts input of keywords at the time of registration,
The document related information storage unit further stores the registration keyword in association with the document,
The keyword search unit calculates a keyword-related frequency at registration based on the keyword at registration,
The document revision management system according to any one of claims 5 to 9, wherein the candidate extraction unit extracts the candidate document related to the registration keyword based on the registration keyword relevance frequency. The keyword search unit calculates, for each of the documents, a registration keyword weight for each of the registration keywords based on the content of the document,
The document related information storage unit stores the registration keyword and the registration keyword weight in association with each other,
The said keyword search part calculates the said keyword relevance degree at the time of registration based on the said keyword weight at the time of registration with respect to each of the said keyword at the time of registration of each of the said documents except the said document for revision. Document revision management system.   The document revision according to claim 11, wherein the keyword search unit calculates the keyword weight at the time of registration based on a keyword highlighting symbol immediately before and a keyword highlighting symbol immediately after the registration keyword appearing in the document. Management system. The keyword search unit, for each of the documents excluding the revision target document, for each of the registration keywords, based on the content of the document, a registration keyword hit indicating the number of times the registration keyword appears. Calculate the number,
The keyword search unit calculates the registration-related keyword relevance frequency based on the registration keyword hit count for each of the registration keywords for each of the documents excluding the revision target document. The document revision management system as described in any one of. The document revision management system further accepts input of a creator ID representing a creator of the document and classification information representing a classification of the document,
The candidate extraction unit includes the document excluding the revision target document,
The document name,
The document revision management system according to any one of claims 5 to 13, wherein a document whose at least one of the creator ID and the classification information matches the revision target document is extracted as a candidate document related to an attribute. The candidate extraction unit further includes a reference analysis unit that identifies a reference name related to the revision target document,
The candidate extraction unit extracts, as the candidate document related to the reference name, a document in which the reference name appears in the document, excluding the revision target document. Document revision management system according to any one of the above. The reference analysis unit specifies a specific character string indicating a boundary between the text of the document to be revised and a reference name list,
The document revision management system according to claim 15, wherein the reference name is specified based on a reference highlighting symbol that appears after a specific character string indicating the boundary.   The reference literature analysis unit identifies the reference name based on a reference highlighting symbol and a specific reference reference character string that appears after the reference highlighting symbol. Document revision management system.   A document revision management program for causing a computer to function as the document revision management system according to any one of claims 1 to 17.

Images