JP2013077174A - Document management device and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a document management device capable of presenting items to be written in a document when a user newly prepares the document, and a program.SOLUTION: Document storage means stores multiple documents in which various types of items are written. Retrieval means retrieves documents matching retrieval criteria specified by a user in the document storage means on the basis of the search retrieval criteria. Acquisition means acquires documents similar to the retrieved documents from the document storage means. Identification means identifies items commonly written in the retrieved document and the acquired documents. Presentation means present the identified items to the user.

Description

  Embodiments described herein relate generally to a document management apparatus and program for presenting items to be described in a document.

  Generally, in a document such as a software specification or a patent, items to be described in the document are often duplicated (that is, common) in similar fields. Specifically, in the case of software specifications, both the Web application that manages employee information and the Web application that manages salary details need to describe specifications (items) related to the screen. Similarly, common items are often described for patent documents.

JP 09-305390 A Japanese Patent Laid-Open No. 08-147152

  By the way, when a user creates a new document as described above, particularly when creating a document related to a field that has not been handled, items to be described in the document may not be known. In such a case, it takes time to create a document, and there are many cases where missing items to be described in the document occur.

  Therefore, the problem to be solved by the present invention is to provide a document management apparatus and program capable of presenting items to be described in a document when a user creates a new document.

  The document management apparatus according to the embodiment includes a document storage unit, a search unit, an acquisition unit, a specification unit, and a presentation unit.

  The document storage means stores a plurality of documents in which various items are described.

  The search means searches the document storage means for a document that matches the search condition based on the search condition specified by the user.

  The acquisition unit acquires a document similar to the searched document from the document storage unit.

  The specifying unit specifies items that are commonly described in the searched document and the acquired document.

  The presenting means presents the identified item to the user.

FIG. 2 is a block diagram illustrating a hardware configuration of the document management apparatus according to the embodiment. FIG. 2 is a block diagram mainly showing a functional configuration of the document management apparatus 30 shown in FIG. 1. The figure for demonstrating notionally the metamodel stored in the document data repository 22 shown in FIG. The figure which shows an example of the data structure of the file of the meta model of an XML format. The block diagram which shows the function structure of the document registration part 32 shown in FIG. The block diagram which shows the function structure of the document search part 33 shown in FIG. The block diagram which shows the function structure of the item acquisition part 34 shown in FIG. 5 is a flowchart showing a processing procedure of document registration processing executed in the document management apparatus 30 according to the present embodiment. 6 is a diagram showing an example of a document describing “functional requirements” input by an input unit 31. FIG. FIG. 6 is a diagram illustrating an example of a document describing “non-functional requirements” input by an input unit 31; The figure for demonstrating concretely about the document converted into the XML format. The figure for demonstrating concretely about the document converted into the XML format. The figure which shows an example of the data structure of the metadata registered into the document data repository. The flowchart which shows the process sequence of a metamodel update process. The figure which shows an example of the data structure of the file of the metamodel after a metamodel update process. 5 is a flowchart showing a processing procedure of item presentation processing executed in the document management apparatus 30 according to the present embodiment. The figure for demonstrating notionally the metamodel in which each item contained in the metadata shown in FIG. 13 was defined as a class. The figure which shows an example of the metadata of the selection document Ds. The figure which shows an example of the metadata of the object document Do. The figure which shows an example of the item shown to a user.

  Hereinafter, embodiments will be described with reference to the drawings.

  FIG. 1 is a block diagram showing a hardware configuration of the document management apparatus according to the present embodiment. As shown in FIG. 1, the computer 10 is connected to an external storage device 20 such as a hard disk drive (HDD). The external storage device 20 stores a program 21 executed by the computer 10. The computer 10 and the external storage device 20 constitute a document management device 30.

  FIG. 2 is a block diagram mainly showing a functional configuration of the document management apparatus 30 shown in FIG. As shown in FIG. 2, the document management apparatus 30 includes an input unit 31, a document registration unit 32, a document search unit 33, an item acquisition unit 34, and an output unit 35. In the present embodiment, these units 31 to 35 are realized by the computer 10 illustrated in FIG. 1 executing the program 21 stored in the external storage device 20. This program 21 can be stored in advance in a computer-readable storage medium and distributed. Further, this program 21 may be downloaded to the computer 10 via, for example, a network.

  The document management apparatus 30 includes a document data repository 22. In the present embodiment, the document data repository 22 is stored in, for example, the external storage device 20.

  The document data repository 22 stores a plurality of documents in which various items are described. Each of the documents stored in the document data repository 22 includes a word.

  The document data repository 22 stores metadata (hereinafter referred to as document metadata) in which detailed information about the document is described for each document stored in the document data repository 22. This metadata includes, for example, items described in the document and words (information indicating) included in the document.

  The document data repository 22 stores items described in each document stored in the document data repository 22 and a model (hereinafter referred to as a meta model) in which the relationship between the items is defined. The

  The input unit 31 inputs a document created by a user (document creator), for example. In the document input by the input unit 31, various items are described as described above, and a word is included.

  Further, the input unit 31 inputs search conditions (for example, search keywords) designated by a user who wants to create a document.

  The document registration unit 32 updates the metamodel stored in the document data repository 22 based on the items described in the document input by the input unit 31. The document registration unit 32 registers the document input by the input unit 31 in the document data repository 22. Further, the document registration unit 32 registers the metadata of the document input by the input unit 31 in the document data repository 22.

  Based on the search condition input by the input unit 31, the document search unit 33 searches the document data repository 22 for a document that matches the search condition.

  The item acquisition unit 34 selects a document similar to the document searched by the document search unit 33 (hereinafter referred to as a similar document) based on the metadata and metamodel of each document stored in the document data repository 22. Obtained from the document data repository 22.

  The item acquisition unit 34 specifies items commonly described in the document searched by the document search unit 33 and a document similar to the document.

  The output unit 35 outputs the item specified by the item acquisition unit 34. Thereby, items to be described in the document are presented to the user.

  Here, the metamodel stored in the document data repository 22 shown in FIG. 2 will be conceptually described with reference to FIG. The document data repository 22 stores one metamodel.

  FIG. 3 shows an example of a meta model related to (a requirement for) software specifications. In the meta model, as described above, items described in each document stored in the document data repository 22 and a relationship between the items are defined. In this metamodel, items described in each document are referred to as classes.

  Software specifications can be created by selecting classes defined in the metamodel. For example, if it is a specification regarding a Web application, a “screen” class (not shown) is selected, and a specification regarding the “screen” is described. On the other hand, if the specifications are related to the framework, the “screen” class is not necessary, so the specifications regarding the “screen” are not described. In addition, if there are items that should be described outside the class defined in the metamodel, they can be added.

  In the meta model shown in FIG. 3, a “functional requirement” class and a “non-functional requirement” class are defined.

  A “common function” subclass and a “function” subclass are defined as subordinate classes (hereinafter referred to as subclasses) of the “functional requirement” class. A “function ID” sub-subclass and a “function name” sub-subclass are defined as subordinate classes (hereinafter referred to as sub-subclasses) of the “function” sub-class. According to this, for example, in a document (software specification) in which “function requirements” are described, items of “common function” and “function” are described, and “function ID” and “function” are attributes as the “function”. It is shown that the item “function name” is described.

  In addition, a “business characteristic” subclass and a “restrictions” subclass are defined as subclasses below the “non-functional requirement” class. As a sub-subclass lower than the “restriction” subclass, a “restriction ID” sub-subclass and a “restriction name” sub-subclass are defined. According to this, for example, in a document (software specification) in which “non-functional requirements” are described, items of “business characteristics” and “restrictions” are described, and “restrictions” are set as attributes of the “restrictions”. It is shown that items of “ID” and “Restriction item name” are described.

  In other words, in the metamodel stored in the document data repository 22, it is described in one document for each node (here, “functional requirement” class and “non-functional requirement” class) of the first hierarchy in the metamodel. Items and their relationships are defined.

  Although the meta model is conceptually described in FIG. 3, the meta model is, for example, an XML (Extensible Markup Language) format file (hereinafter referred to as a meta model file) as shown in FIG. 22.

  Note that the metamodel file shown in FIG. 4 includes a “class” element, a “subclass” element, and a “subclass” element that define each class. Each of the “class” element, the “subclass” element, and the “subclass” element has a “name” attribute whose attribute value is the class name (that is, item name) of the class defined by the element.

  In addition, the “class” element, the “subclass” element, and the “subclass” element in the metamodel file shown in FIG. 4 constitute a hierarchical structure of each class defined in the metamodel. Here, the “subclass” element defines a subclass lower than the class defined in the “class” element. Further, the “subsubclass” element defines a sub-subclass lower than the subclass defined in the “subclass” element. Each class defined in the meta model is given an identifier (id) for identifying the class as an attribute value. Specifically, the “class” element, the “subclass” element, and the “subclass” element have an “id” attribute whose attribute value is an identifier for identifying a class defined by the element.

  FIG. 5 is a block diagram showing a functional configuration of the document registration unit 32 shown in FIG. As shown in FIG. 5, the document registration unit 32 includes a document acquisition unit 321, a metamodel acquisition unit 322, a metamodel update unit 323, a metadata generation unit 324, and a registration unit 325.

  The document acquisition unit 321 acquires the document input by the input unit 31. The document acquisition unit 321 converts the acquired document into, for example, an XML format.

  The meta model acquisition unit 322 acquires the meta model (file) from the document data repository 22. This meta model is a model in which the items described in each of the documents stored in the document data repository 22 and the relationship between the items are defined as described above.

  The metamodel update unit 323 compares the items described in the document converted into the XML format by the document acquisition unit 321 and the classes defined in the metamodel, thereby comparing the items described in the document. To extract items (classes) that are not defined in the metamodel. The metamodel update unit 323 updates the metamodel by adding (appending) the extracted class to the metamodel file.

  The metadata generation unit 324 generates metadata of the document based on the document converted into the XML format by the document acquisition unit 321. In this case, the metadata generation unit 324 includes the items described in the document converted into the XML format by the document acquisition unit 321, the relationship between the items, the words included in the document, and the number of occurrences of the word in the document ( Metadata including information) is generated.

  The registration unit 325 registers the document converted into the XML format by the document acquisition unit 321 and the metadata generated by the metadata generation unit 324 in the document data repository 22.

  FIG. 6 is a block diagram showing a functional configuration of the document search unit 33 shown in FIG. As shown in FIG. 6, the document search unit 33 includes a search condition acquisition unit 331 and a search result acquisition unit 332.

  The search condition acquisition unit 331 acquires the search condition input by the input unit 31. This search condition includes, for example, a search keyword specified by the user.

  The search result acquisition unit 332 searches the document data repository 22 for a document that matches the search condition based on the search condition (search keyword) acquired by the search condition acquisition unit 331. As a result, the search result acquisition unit 332 acquires a search result including a document that matches the search condition acquired by the search condition acquisition unit 331. Note that the search result acquired by the search result acquisition unit 332 is presented to the user via the output unit 35.

  FIG. 7 is a block diagram showing a functional configuration of the item acquisition unit 34 shown in FIG. As illustrated in FIG. 7, the item acquisition unit 34 includes a selected document acquisition unit 341, a metadata acquisition unit 342, a similar document acquisition unit 343, and an item calculation unit 344.

  The selected document acquisition unit 341 acquires the document selected by the user from the search results (documents included) presented to the user.

  The metadata acquisition unit 342 acquires the metadata of all documents stored in the document data repository 22 from the document data repository 22.

  The similar document acquisition unit 343 calculates the similarity between the document acquired by the selected document acquisition unit 341 and the document stored in the document data repository 22 based on the metadata acquired by the metadata acquisition unit 342. The similar document acquisition unit 343 acquires a document similar to the document acquired by the selected document acquisition unit 341 based on the calculated similarity.

  The item calculation unit 344 acquires a meta model from the document data repository 22. The item calculation unit 344 describes, for each class (item) defined in the acquired metamodel, the item in the document acquired by the selected document acquisition unit 341 and the document acquired by the similar document acquisition unit 343. The number of documents being calculated is calculated as the degree of duplication. The item calculation unit 344 specifies items commonly described in the document acquired by the selected document acquisition unit 341 and the document acquired by the similar document acquisition unit 343 based on the calculated degree of duplication. The item specified by the item calculation unit 344 is presented to the user via the output unit 35 as an item to be described in the document that the user intends to create.

  Hereinafter, the operation of the document management apparatus 30 according to the present embodiment will be described. In the document management apparatus 30 according to the present embodiment, a process for registering a document created by a user in the document data repository 22 (hereinafter referred to as document registration process) and a new document created by the user are included in the document. Processing for presenting items to be described (hereinafter referred to as item presentation processing) is executed.

  Here, with reference to the flowchart of FIG. 8, a processing procedure of document registration processing executed in the document management apparatus 30 according to the present embodiment will be described. Here, it is assumed that the document data repository 22 stores a metamodel file as shown in FIG.

  First, the input unit 31 inputs a document created by the user in accordance with, for example, a user operation (step S1). Various items are described in the document input by the input unit 31. The document input by the input unit 31 includes words. Note that a plurality of documents may be input by the input unit 31.

  The document input by the input unit 31 may be any type of document such as a software specification, patent specification, or regulation document, but is based on a metamodel stored in the document data repository 22. Document.

  Here, with reference to FIG. 9 and FIG. 10, the document input by the input part 31 is demonstrated concretely.

  FIG. 9 shows an example of a document in which “functional requirements” are described in the software specification. In the example illustrated in FIG. 9, a document in which “functional requirements” are described includes sheets of “common function”, “function”, “business logic”, “exception processing”, and “screen”. In the sheet, “function ID”, “function name”, “description”, and “detailed description” are described. Note that “function ID”, “function name”, “description”, and “detailed description” described in the “function” sheet are attributes of the “function”.

  In the document shown in FIG. 9, “functional requirements”, “common functions”, “functions”, “business logic”, “exception processing”, “screen”, “function ID”, “function name”, “explanation”. And “detailed description” are items described in the document.

  On the other hand, FIG. 10 shows an example of a document in which “non-functional requirements” are described in the software specification. In the example illustrated in FIG. 10, a document in which “non-functional requirements” is described includes sheets of “business characteristics”, “restrictions”, “memory constraints”, “design constraints”, and “working locations”. In the “restriction item” sheet, “restriction item ID”, “restriction item name”, “explanation”, and “origin” are described. Note that “restriction ID”, “restriction name”, “description”, and “derived” described in the “restriction” sheet are attributes of the “restriction”.

  In the document shown in FIG. 10, “non-functional requirements”, “business characteristics”, “restrictions”, “memory constraints”, “design constraints”, “operation locations”, “restriction IDs”, “restriction names” ”,“ Description ”, and“ Origin ”are items described in the document.

  Returning to FIG. 8 again, the document acquisition unit 321 included in the document registration unit 32 acquires the document input by the input unit 31. Note that the document acquired by the document acquisition unit 321 is converted into the XML format (step S2). When there are a plurality of documents acquired by the document acquisition unit 321, all of the plurality of documents are converted into one XML format document (file).

  Here, with reference to FIG. 11 and FIG. 12, the document converted into the XML format will be specifically described. FIGS. 11 and 12 show examples in which the documents shown in FIGS. 9 and 10 are converted into XML format files.

  In the example shown in FIGS. 11 and 12, the content of the “functional requirement” element (that is, the content from <functional requirement> to </ functional requirement>) corresponds to the document in which “functional requirement” shown in FIG. 9 is described. To do. The “functional requirement” element includes an element indicating each item described in the document in which the “functional requirement” is described. Further, the content of the element indicating each item described in the document in which “functional requirements” is described indicates the content of the document described in the item.

  Similarly, in the examples shown in FIGS. 11 and 12, the contents of the “non-functional requirements” element (contents from <non-functional requirements> to </ non-functional requirements>) describe “non-functional requirements” shown in FIG. Corresponds to the document. The “non-functional requirement” element includes an element indicating each item described in the document in which “non-functional requirement” is described. Further, the content of the element indicating each item described in the document in which “non-functional requirement” is described indicates the content of the document described in the item.

  Returning to FIG. 8 again, the metamodel acquisition unit 322 acquires a metamodel (file) from the document data repository 22.

  The metamodel update unit 323 executes metamodel update processing based on the document acquired by the document acquisition 321 (the document converted into the XML format) and the metamodel acquired by the metamodel acquisition unit 322 (step S3). ). According to this metamodel update processing, items that are not defined in the metamodel are extracted from items described in the document acquired by the document acquisition unit 321, and the extracted items are included in the metamodel as classes. By adding (adding), the metamodel is updated. Details of the metamodel update process will be described later.

  Next, the metadata generation unit 324 performs morphological analysis on the content (character string) of each item described in the document acquired by the document acquisition unit 321. Specifically, a character string included in the content of an element indicating each item described in the document converted into the XML format is subjected to morphological analysis. The metadata generation unit 324 acquires words included in the document acquired by the document acquisition unit 321 based on the morphological analysis result.

  The metadata generation unit 324 generates metadata of the document based on the document acquired by the document acquisition unit 321 and the words included in the document (step S4). Note that the metadata generated by the metadata generation unit 324 includes items described in the document acquired by the document acquisition unit 321, relationships between the items, words included in the document, and the words in the document. The number of occurrences (number of occurrences) is included.

  The registration unit 325 registers (stores) the document acquired by the document acquisition unit 321 and the metadata of the document generated by the metadata generation unit 324 in the document data repository 22 (step S5). The registration unit 325 adds the registration date (information indicating) to the metadata when registering the metadata in the document data repository 22. That is, the metadata of the document registered in the document data repository 22 includes the registration date of the document (and metadata).

  Here, FIG. 13 shows an example of the data structure of the metadata registered in the document data repository 22. FIG. 13 shows metadata of a document identified by doc — 00001, for example.

  According to the metadata shown in FIG. 13, the registration date of the metadata is indicated in the “date” element. In addition, each item described in the document included in the metadata is indicated in each element included in the “metamodel” element. Further, the word included in the XML document included in the metadata and the number of occurrences of the word are indicated as attribute values of the “name” attribute and the “num” attribute included in the “word” element included in the “words” element. Yes.

  As shown in FIG. 13, the metadata registered in the document data repository 22 is described in the XML format.

  Next, the processing procedure of the metamodel update process (the process of step S3 shown in FIG. 8) will be described with reference to the flowchart of FIG. According to this metamodel update process, items (classes) not defined in the metamodel among items described in the document acquired by the document acquisition unit 321 included in the document registration unit 32 are added to the metamodel. Is done. This metamodel update process is executed by the metamodel update unit 323 included in the document registration unit 32.

  In the metamodel update process, first, the following steps S11 to S15 are executed for each element of the first hierarchy in the document acquired by the document acquisition unit 321 (the document converted into the XML format).

  In this case, the metamodel update unit 323 acquires one element of the first layer in the document acquired by the document acquisition unit 321 (the document converted into the XML format) (Step S11). Here, the first layer element in the document acquired by the document acquisition unit 321 refers to an element one layer lower than the root element in the document. More specifically, referring to the documents shown in FIG. 11 and FIG. 12, the “functional requirement” element and the “non-functional requirement” element correspond to the elements of the first hierarchy in the document. Hereinafter, the element acquired in step S11 will be described as a target element.

  Next, the metamodel update unit 323 selects an element (hereinafter referred to as a corresponding element) corresponding to the target element (the first layer element acquired in step S11) from the metamodel file stored in the document data repository 22. (Notation) is acquired (step S12). Here, the “corresponding element” refers to an element in the metamodel file in the same hierarchy as the element hierarchy acquired in step S11. The metamodel update unit 323 acquires the attribute value of the “name” attribute of the corresponding element.

  More specifically, with reference to the metamodel file shown in FIG. 4, when the target element is an element in the first hierarchy, the corresponding element is an element in the first hierarchy in the metamodel, for example, “class (class)” Element. In this case, the metamodel update unit 323 acquires the attribute value “functional requirement” of the “name” attribute of the “class” element that is the corresponding element. Hereinafter, the attribute value of the “name” attribute of the corresponding element of the meta model is simply referred to as the attribute value of the corresponding element.

  The metamodel update unit 323 determines whether or not the target element (the element name) acquired as described above matches the attribute value of the corresponding element (step S13).

  When it is determined that the element name of the target element and the attribute value of the corresponding element do not match (NO in step S13), the metamodel update unit 323 stores in the document data repository 22 whether or not there is another corresponding element. The determination is made with reference to the metamodel (file) (step S14). That is, the metamodel update unit 323 determines whether there is a corresponding element other than the corresponding element acquired in step S12.

  When it is determined that there is a corresponding element of another metamodel (YES in step S14), the process returns to the above-described step S12 and is repeated.

  On the other hand, when it is determined that there is no corresponding element of another meta model (NO in step S14), the meta model update unit 323 sets the target element as a class to a meta model file stored in the document data repository 22. Add (append) (step S15). Specifically, the metamodel update unit 323 adds an element having a “name” attribute (here, “class” element) having the element name of the target element as an attribute value to the metamodel file. As a result, the metamodel (file) stored in the document data repository 22 is updated. Here, in the document acquired by the document acquisition unit 321, when an element exists in a hierarchy lower than the target element, all of the target element and the elements lower in the hierarchy are included in the metamodel file. Added.

  When the process of step S15 is executed, it is determined whether or not the processes of steps S11 to S15 described above have been executed for all the first layer elements in the document acquired by the document acquisition unit 321 (step S16). .

  When it is determined that the process is not executed for all the elements of the first hierarchy, the process returns to the above-described step S11 and the process is repeated.

  On the other hand, when it is determined that the process has been executed for all the elements of the first hierarchy, the metamodel update process for the elements of the first hierarchy is terminated.

  Here, the metamodel update processing for the first layer element in the document acquired by the document acquisition unit 321 (the document converted into the XML format) has been described, but the same applies to the second layer element in the document. Metamodel update processing is executed.

  Of the elements in the second hierarchy, the elements to be subjected to the metamodel update process are elements below the elements in the first hierarchy determined to match the corresponding elements in step S13 described above. That is, since the elements below the first hierarchy element determined not to match the corresponding element in step S13 have already been added to the metamodel in step S15, the metamodel update process is not executed.

  In this case, the process is executed with the element of the first hierarchy in the metamodel update process described above as the element of the second hierarchy. The same applies to elements in the third and subsequent layers.

  By executing the metamodel update process in this way, whether each element of each hierarchy in the document acquired by the document acquisition unit 321 is defined as a class in the metamodel stored in the document data repository 22 It is determined whether or not. As a result, elements that are not defined as classes in the metamodel (and elements that are in a lower hierarchy of the element) are added to the metamodel.

  FIG. 15 shows an example of the data structure of the metamodel file after the metamodel update process. Here, it is assumed that the metamodel file shown in FIG. 4 has been updated.

  In the meta model file shown in FIG. 15, a “subclass” element having an attribute value “detailed description” is added as compared to the meta model file shown in FIG.

  According to this, the “detailed description” element (that is, the item of “detailed description” described in the document) in the document acquired by the document acquisition unit 321 is not defined in the metamodel as a class, It is shown that the “detailed description” element has been added to the metamodel.

  Next, with reference to a flowchart of FIG. 16, a processing procedure of item presentation processing executed in the document management apparatus 30 according to the present embodiment will be described. Here, in the document data repository 22, as described above, one metamodel in which a plurality of documents, metadata for each document, items described in each of the plurality of documents, and relationships between the items are defined. Is stored. The documents, metadata, and metamodel stored in the document data repository 22 are XML format files as described above.

  First, the input unit 31 inputs a search keyword designated by the user (step S21).

  Next, the search condition acquisition unit 331 included in the document search unit 33 acquires the search keyword input by the input unit 31.

  The search result acquisition unit 332 included in the document search unit 33 uses the search keyword acquired by the search condition acquisition unit 331 to search the document data repository 22 for a document that matches the search keyword (step S22). As a result, the search result acquisition unit 332 acquires a search result including a document that matches the search keyword.

  The search result acquisition unit 332 outputs the acquired search result via the output unit 35 (step S23). As a result, the search results (documents included) are presented to the user.

  Here, the user can select a desired document for the user from the presented search results. Specifically, the user can select, for example, a document similar to a document that the user intends to create.

  In the document management apparatus 30, it is determined whether or not a document has been selected by the user (step S24).

  If it is determined that the user has not selected a document (NO in step S24), the process returns to step S21 and the process is repeated.

  On the other hand, when it is determined that a document has been selected by the user (YES in step S24), the selected document acquisition unit 341 included in the item acquisition unit 34 displays the document selected by the user (hereinafter referred to as a selected document). For example, it is acquired from the document data repository 22 (step S25).

  Next, the metadata acquisition unit 342 included in the item acquisition unit 34 acquires metadata of all documents stored in the document data repository 22 (step S26).

  The similar document acquisition unit 343 included in the item acquisition unit 34 stores a document similar to the selected document (hereinafter referred to as a similar document) in the document data repository 22 based on the metadata acquired by the metadata acquisition unit 342. It is obtained (calculated) from the document being processed (step S27).

  In this case, the similar document acquisition unit 343 calculates the similarity between the selected document and each of the documents stored in the document data repository 22. The similar document acquisition unit 343 acquires a document whose calculated similarity is equal to or greater than a predetermined value (threshold value) as a similar document (a document similar to the selected document).

  Here, the similarity calculation processing by the similar document acquisition unit 343 will be described. Here, it is assumed that the similar document acquisition unit 343 calculates the first and second similarities. In the following description, a document whose similarity with the selected document is calculated is a target document.

  First, the first similarity calculated by the similar document acquisition unit 343 will be described. The first similarity is a meta model in which each item (items described in the selected document and the relationship between the items) included in the metadata of the selected document is defined as a class (hereinafter referred to as a meta model of the selected document). Notation) and the meta model (hereinafter referred to as the meta model of the target document) in which each item (the item described in the target document and the relationship between the items) included in the metadata of the target document is defined as a class Is calculated.

  Here, a meta model of a document (selected document and target document) will be specifically described with reference to FIG. The meta model of the document is generated based on the metadata of the document. A document metamodel is generated for each document stored in the document data repository 22.

  In the example illustrated in FIG. 17, the metamodel of the document includes “metamodel” element of the metadata of the document and each element (item indicated by) included in the “metamodel” element as a class.

  Specifically, the meta model of a document includes a “class” element (element that is one layer lower than the “metamodel” element) and a “subclass” element (element that is one layer lower than the “class” element) of the metadata of the document. ) And “subsubclass” elements (elements one level lower than the “subclass” element) are included as classes.

  In the calculation of the first similarity, a comparison is made with the target document in units of subtrees composed of elements (classes) in the first hierarchy and below when the meta model of the selected document is a tree structure. In the example shown in FIG. 17, a portion surrounded by broken lines 100a and 100b is a subtree to be compared with the target document.

  Here, the meta model of the selected document is Ms, and the meta model of the target document is Mo. If the similarity between Ms and Mo (that is, the first similarity) is sim_metamodel (Ms, Mo), sim_metamodel (Ms, Mo) is expressed by the following equation (1).

  Msi in Expression (1) represents each subtree of Ms. Class (Msi) represents the number of classes included in the subtree of Ms (that is, Msi). Further, match (Mo) represents the number of classes included in Mo that matches the class included in Msi.

  Next, the second similarity calculated by the similar document acquisition unit 343 will be described. The second similarity is calculated using words (words included in the document) included in the metadata of the selected document and the target document and the number of appearances of the word.

  Here, tf / idf values of words included in the selected document and the target document are calculated. Here, the number of occurrences (frequency) of the word t in the document d is assumed to be tf (t, d). The word t is defined in the “word” element included in the metadata of the document d. Tf (t, d) is an attribute value of the “num” attribute of the “word” element included in the metadata of the document d.

  If the number of documents stored in the document data repository 22 is N and the number of documents in which the word t appears one or more times is df (t), idf (t) is expressed by the following equation (2).

  Further, the weight w (t, d) of the word t in the document d is expressed by the following equation (3).

Furthermore, if the selected document is Ds and the tf / idf values of each word included in the selected document Ds are w ₁ , w ₂ ,..., W _p , the vector of the selected document Ds is expressed by the following equation (4) Is done.

If the target document is Do and the tf / idf values of each word included in the target document Do are w ₁ , w ₂ ,..., W _p , the vector of the target document Do is expressed by the following equation (5). Is done.

  The vector of the target document Do includes the tf / idf value of each word corresponding to the tf / idf value included in the vector of the selected document Ds. That is, when the target document Do does not include a word included in the selected document Ds, the tf / idf value of the word included in the target document Do is 0.

  Assuming that the similarity (that is, the second similarity) using the words included in the selected document Ds and the target document Do and the number of occurrences of the words is sim_words (Ds, Do), sim_words (Ds, Do) is as follows: It is represented by Formula (6).

  As described above, when the first and second similarities are calculated, the similar document acquisition unit 343 calculates the similarity between the selected document and the target document. When the similarity between the selected document Ds and the target document Do is sim (Ds, Do), sim (Ds, Do) is expressed by the following equation (7).

  When the similarity between the selected document Ds calculated by Expression (7) and the target document Do is equal to or greater than the threshold, the similar document acquisition unit 343 determines that the target document Do is similar to the selected document Ds, that is, a similar document. Get as.

  Hereinafter, the calculation process of the similarity (that is, sim (Ds, Do)) between the selected document Ds and the target document Do will be specifically described. Here, it is assumed that the metadata shown in FIG. 18 is the metadata of the selected document Ds. Further, it is assumed that the metadata shown in FIG. 19 is the metadata of the target document Do.

  First, the first similarity calculated using the meta model of the selected document Ds and the target document Do will be specifically described. Here, the meta model of the selected document Ds is Ms, and the meta model of the target document Do is Mo.

  Here, according to the metadata of the selected document Ds shown in FIG. 18, the metamodel Ms of the selected document Ds is a subtree having a “class” element whose attribute value of “id” attribute is “meta — 001” as a root element. (Hereinafter referred to as a partial tree Ms1) and a partial tree (hereinafter referred to as a partial tree Ms2) having a “class” element having an attribute value of “meta_002” as a root element.

  In the subtree Ms1, the “class” element whose attribute value of the “id” attribute included in the metadata of the selected document Ds is “meta — 001”, the “subclass” element and the “subclass” element subordinate to the “class” element. The element (item indicated by) is included as a class. Specifically, according to the metadata of the selected document Ds illustrated in FIG. 18, the subtree Ms1 includes a “class” element having an attribute value of “meta_001” and an attribute value of the “id” attribute in the subtree Ms1. Is a “subclass” element whose meta attribute is “meta — 001 — 001”, an “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0001”, and a “subclass” element whose attribute value is “meta — 001 — 0001 — 0002” An “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0003” and an “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0007” are included. Therefore, class (Ms1) representing the number of classes (elements) included in the subtree Ms1 is 6.

  Further, according to the metadata of the target document Do shown in FIG. 19, the meta model Mo of the target document Do includes the “class” element whose attribute value of the “id” attribute is “meta — 001”, and the attribute of the “id” attribute. A “subclass” element whose value is “meta — 001 — 0001”, an “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0001”, and a “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0002” The “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0004”, the “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0009”, and the attribute value of the “id” attribute is “me” The “class” element with “a — 002”, the “subclass” element with the attribute value of “id” attribute “meta — 002 — 0001”, the “subsubclass” element with the attribute value of “meta — 002 — 0001 — 0001”, and the “id” attribute A “subsubclass” element whose attribute value is “meta — 002 — 0001 — 0002” and an “subsubclass” element whose “id” attribute attribute value is “meta — 002 — 0001 — 0005” (items indicated by) are included as classes.

  In this case, the class (item) included in the subtree Ms1 that matches the class included in the meta model Mo of the target document Do has the “class” element whose attribute value “meta — 001” is “meta_001”, “id”. The “subclass” element whose attribute value is “meta — 001 — 0001”, the “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0001”, and the “subclass” element whose attribute value is “meta — 001 — 0001 — 0002” It is. Therefore, match (Mo) representing the number of classes included in the subtree Ms1 that matches the class included in the meta model Mo of the target document Do is 4.

  On the other hand, in the subtree Ms2, the “class” element whose attribute value of the “id” attribute included in the metadata of the selected document Ds is “meta — 002”, the “subclass” element subordinate to the “class” element, and “ The “subclass” element is included as a class. Specifically, according to the metadata of the selected document Ds shown in FIG. 18, in the subtree Ms2, the “class” element whose attribute value of the “id” attribute is “meta — 002”, and the attribute value of the “id” attribute "Subclass" element whose attribute value is "meta_002_0001", "subsubclass" element whose attribute value is "meta_002_0001_0001", and "subclass" element whose attribute value is "meta_002_0001_0002" and "id" An “subsubclass” element whose attribute value is “meta — 002 — 0001 — 0003” is included. Therefore, class (Ms2) representing the number of classes included in the subtree Ms2 is 5.

  In this case, the class (item) included in the subtree Ms2 that matches the class included in the meta model Mo of the target document Do described above has a “class” element whose attribute value of “id” attribute is “meta — 002”, “ The “subclass” element whose attribute value is “meta — 002 — 0001”, the “subsubclass” element whose attribute value is “meta — 002 — 0001 — 0001”, and the “subsubclass” whose attribute value is “meta — 002 — 0001 — 0002” Element. Therefore, match (Mo) representing the number of classes included in the subtree Ms2 that matches the class included in the meta model Mo of the target document Do is 4.

  Using the above values, the similarity (first similarity) sim_metamodel (Ms, Mo) between the meta model Ms of the selected document Ds and the meta model Mo of the target document Do is 4/6 + 4/5 = 1.3. Is calculated.

  Next, the second similarity calculated using the words included in the metadata of the selected document Ds and the target document Do and the number of appearances of the words will be specifically described. Here, the number N of documents stored in the document data repository 22 is assumed to be 10.

  According to the metadata of the selected document Ds shown in FIG. 18, the selected document Ds includes the words “customer”, “information”, “management”, and “search”. The number of words “customer” included in the selected document Ds (that is, the number of appearances) is 10. The number of words “information” included in the selected document Ds is 20. The number of words “management” included in the selected document Ds is five. The number of words “search” included in the selected document Ds is five.

  The word included in the selected document Ds can be obtained from the attribute value of the “name” attribute of the “word” element of the metadata of the selected document Ds. Further, the number of appearances of words included in the selected document Ds can be obtained from the attribute value of the “num” attribute included in the “word” element of the metadata of the selected document Ds.

  Here, among the documents stored in the document data repository 22, the number of documents in which the word “customer” included in the selected document Ds appears at least once (that is, df (customer)) is 3. To do. In this case, the tf / idf value (that is, w (customer, selected document Ds)) of the word “customer” included in the selected document Ds is 10 * log (10/3) = 5.2.

  The number of documents in which the word “information” included in the selected document Ds appears one or more times (ie, df (information)) among the documents stored in the document data repository 22 is 5. . In this case, the tf / idf value (that is, w (information, selected document Ds)) of the word “information” included in the selected document Ds is 20 * log (10/5) = 6.

  In addition, among the documents stored in the document data repository 22, the number of documents in which the word “management” included in the selected document Ds appears one or more times (that is, df (management)) is 1. . In this case, the tf / idf value (that is, w (management, selected document Ds)) of the word “management” included in the selected document Ds is 5 * log (10/1) = 5.

  The number of documents in which the word “search” included in the selected document Ds appears at least once among documents stored in the document data repository 22 (that is, df (search)) is 3. . In this case, the tf / idf value (that is, w (search, selected document Ds)) of the word “search” included in the selected document Ds is 5 * log (10/3) = 2.6.

  According to this, the vector of the selected document Ds is expressed by the following equation (8).

  On the other hand, according to the metadata of the target document Do shown in FIG. 19, the target document Do includes the words “employee”, “information”, “search”, and “registration”. The number of words “employee” (that is, the number of appearances) included in the target document Do is 10. The number of words “information” included in the target document Do is 19. The number of words “search” included in the target document Do is six. Further, the number of words “registration” included in the target document Do is six.

  The word included in the target document Do can be obtained from the attribute value of the “name” attribute of the “word” element of the metadata of the target document Do. Further, the number of appearances of words included in the target document Do can be obtained from the attribute value of the “num” attribute included in the “word” element of the metadata of the target document Do.

  Here, among documents stored in the document data repository 22, the number of documents in which the word “employee” included in the target document Do appears one or more times (that is, df (employee)) is 1. Shall. In this case, the tf / idf value (that is, w (employee, target document Do)) of the word “employee” included in the target document Do is 10 * log (10/1) = 10.

  The number of documents in which the word “information” included in the target document Do appears one or more times (ie, df (information)) among the documents stored in the document data repository 22 is 5. . In this case, the tf / idf value (that is, w (information, target document Do)) of the word “information” included in the target document Do is 19 * log (10/5) = 5.7.

  The number of documents in which the word “search” included in the target document Do appears one or more times (ie, df (search)) among documents stored in the document data repository 22 is 3. . In this case, the tf / idf value (that is, w (search, target document Do)) of the word “search” included in the target document Do is 6 * log (10/3) = 3.1.

  The number of documents in which the word “registration” included in the target document Do appears one or more times (ie, df (registration)) among the documents stored in the document data repository 22 is 3. . In this case, the tf / idf value (that is, w (registration, target document Do)) of the word “registration” included in the target document Do is 6 * log (10/3) = 3.1.

  Here, as described above, the vector of the target document Do includes the tf / idf value of each word corresponding to the tf / idf value included in the vector of the selected document Ds. According to this, the vector of the target document Do is as shown in the following equation (9).

  Using the vector of the selected document Ds shown in the equation (8) and the vector of the target document Do shown in the equation (9), the similarity using the words included in the selected document Ds and the target document Do and the number of occurrences of the word is used. The degree (second similarity) sim_words (Ds, Do) can be calculated as in the following Expression (10).

  When the above-described sim_metamodel (Ms, Mo) and sim_words (Ds, Do) are used, the similarity sim (Ds, Do) between the selected document Ds and the target document Do is sim_metamodel (Ms, Mo) + sim_words (Ds, Do). = 1.3 + 0.67 = 1.97. If the threshold for acquiring a similar document is 1.5, for example, the target document Do is acquired as a document similar to the selected document Ds.

  Although the similarity between the selected document Ds and the target document Do has been described here, the similarity with the selected document Ds is calculated for all the documents stored in the document data repository 22.

  Returning to FIG. 16 again, the item calculation unit 344 included in the item acquisition unit 34 uses items (classes) defined in the metamodel stored in the document data repository 22 using the selected document and the similar document. The degree of duplication is calculated every time (step S28). The degree of duplication indicates the number of selected documents and similar documents in which items defined in the metamodel appear (that is, the number of duplicates).

  Here, the degree of duplication calculated by the item calculation unit 344 will be described. Here, it is assumed that the selected document is document A and the similar documents are documents B to E.

  In this case, the item calculation unit 344 determines whether an item (class) defined in the meta model is included in each of the documents A to E. Specifically, the item calculation unit 344 determines whether or not the “class” element, the “subclass” element, the “subsubclass” element, and the like defined in the metamodel are included in the metadata of the documents A to E. To do. This determination processing is performed by the attribute value of the “id” attribute included in the “class” element, the “subclass” element, and the “subclass” element defined in the meta model, and the “class” element included in the metadata of the documents A to E. , “Subclass” element and “subsubclass” element are compared with the attribute value of the “id” attribute.

  If a certain item defined in the meta model is included in all of the documents A to E, the duplication degree of the item is 5. On the other hand, when a certain item defined in the meta model is not included in all of the documents A to E, the duplication degree of the item is zero.

  Next, the item calculation unit 344 specifies an item (class) whose calculated degree of overlap is a predetermined value (threshold value) or more.

  The item calculation unit 344 outputs the identified item via the output unit 35 (step S29). As a result, the item specified by the item calculation unit 344 (that is, the item whose degree of overlap is greater than or equal to the threshold) is presented to the user as an item to be described in the document.

  Here, FIG. 20 shows an example of items presented to the user. In the example illustrated in FIG. 20, items presented to the user include a “class” element whose attribute value of the “id” attribute is “meta — 001”, and “subclass” whose attribute value of the “id” attribute is “meta — 001 — 0001”. The “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0001”, the “subsubclass” element whose attribute value is “meta — 001 — 0001 — 0002”, and the attribute value of the “id” attribute is “meta — 002” A certain “class” element, a “subclass” element whose attribute value of “id” attribute is “meta — 002 — 0001”, a “subsubclass” element whose attribute value of “id” attribute is “meta — 002 — 0001 — 0001”, and an “id” genus Value of the attribute is included in the "subsubclass" element is "meta_002_0001_0002".

  The items to be described in the document may be presented to the user in the format shown in FIG. 20, but are stored in the document data repository 22 so that the user can more easily recognize the items to be described. By referring to the meta model, for example, the item name of the item (the “class” attribute attribute value of the “class” element, the “subclass” element, and the “subsubclass” element included in the meta model) is preferably presented. .

  As described above, in the present embodiment, a document that matches the search condition is searched from the document data repository 22 based on the search condition (search keyword) specified by the user, and a document similar to the searched document is searched. An item acquired from the document data repository 22, an item commonly described in the searched document and the acquired document is specified, and the specified item is presented to the user, so that the user can create a new document. At the time of creation, items to be described in the document can be presented. That is, in the present embodiment, by presenting items to be described in a document that the user intends to create from now on, it is possible to reduce the man-hours when the user creates a new document, and also necessary for the document. It is possible to prevent items from being dropped.

  In the present embodiment, a document created by the user is input, the input document is registered in the document data repository 22, and among the items described in the input document, the document data repository 22 is registered. Depending on the items described in the document when the document is created by the user, the metamodel is updated by adding an item that is not defined in the stored metamodel to the metamodel. The metamodel can be automatically updated.

  In the present embodiment, the description has been made assuming that a document (similar document) similar to the selected document is acquired based on the first and second similarities. For example, any one of the first and second similarities is obtained. A configuration may be used in which similar documents are acquired based only on one side. Moreover, the calculation method of the 1st and 2nd similarity mentioned above is an example, and the structure that a similar document is acquired based on the similarity calculated by another method may be sufficient.

  Further, the present invention is not limited to the above-described embodiments as they are, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment.

  DESCRIPTION OF SYMBOLS 10 ... Computer, 20 ... External storage device, 22 ... Document data repository (document storage means), 30 ... Document management apparatus, 31 ... Input part, 32 ... Document registration part, 33 ... Document search part, 34 ... Item acquisition part, 35: Output unit.

Claims (6)

Document storage means for storing a plurality of documents in which various items are described;
A search unit that searches the document storage unit for a document that matches the search condition based on a search condition specified by a user;
Obtaining means for obtaining a document similar to the retrieved document from the document storage means;
Specifying means for specifying items commonly described in the retrieved document and the acquired document;
A document management apparatus comprising: presenting means for presenting the identified item to the user. The acquisition means includes
Generating means for generating, for each document stored in the document storing means, a metamodel in which items described in the document are defined;
Similarity calculation means for calculating the similarity between the retrieved document and each of the documents stored in the document storage means based on the generated metamodel;
The document management apparatus according to claim 1, wherein a document similar to the searched document is acquired from the document storage unit based on the calculated similarity. Each of the documents stored in the document storage means includes a word,
The acquisition means includes
Calculation means for calculating the similarity between the retrieved document and each of the documents stored in the document storage means based on words included in each of the documents stored in the document storage means;
The document management apparatus according to claim 1, wherein a document similar to the searched document is acquired from the document storage unit based on the calculated similarity. The document storage means further stores a meta model in which items described in each of the documents stored in the document storage means and a relationship between the items are defined,
The specifying means is:
For each item defined in the metamodel, including a degree-of-duplication calculation means for calculating the number of the retrieved document and the acquired document in which the item is described as a degree of redundancy,
The document management apparatus according to claim 1, wherein an item commonly described in the retrieved document and the acquired document is specified based on the calculated degree of duplication. An input means for inputting a document created by the user;
Registration means for registering the input document in the document storage means;
Updating means for updating the metamodel by adding items that are not defined in the metamodel stored in the document storage means among the items described in the input document to the metamodel; The document management apparatus according to claim 4, further comprising: A program executed by a computer in a document management apparatus including an external storage device having a document storage means for storing a plurality of documents in which various items are described, and a computer using the external storage device. ,
In the computer,
Searching the document storage means for a document that matches the search condition based on the search condition specified by the user;
Obtaining a document similar to the retrieved document from the document storage means;
Identifying items commonly described in the retrieved document and the acquired document;
A program for causing the identified item to be presented to the user.

Images