JP2009211480A - Structured document processing system, structured document processing method, and structured document processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a structured document processing technology for mapping sections without rendering a structured document and mapping sections considering both a layout and a structure of the structured document. <P>SOLUTION: Weight is assigned to each tag which constitutes a structured document on the basis of the degree of influence to a layout and the degree of non-similarity among structured documents is calculated on the basis of the assigned weight and a tag structure within each structured document. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a structured document processing system, a structured document processing method, and a structured document processing program for processing a structured document, and more particularly to a specific document type definition (DTD: Document Type) such as HTML (Hyper Text Markup Language). A structured document processing system, a structured document processing method, and a structured document processing method for mapping a region divided under a specific condition to another region similar in both layout and structure in a structured document that forms a layout based on (Definition) The present invention relates to a structured document processing program.

  In recent years, a system that automatically analyzes a document structure and extracts a plurality of sections from a structured document such as a Web content including a large amount of information has been studied. For example, there is an information providing system that automatically or manually selects and uses information required by a user from a section. The section is an area obtained by subdividing the structured document, and is configured by a markup language (hereinafter referred to as a tag) as in the structured document.

  However, in applications that use Web content sections whose contents change constantly, such as the information provision system, the display position of the section changes or disappears over time. May provide incorrect section information.

  Therefore, in Patent Document 1, paying attention to the characteristics of the layout of the structured document, and using layout information such as the section extraction order, section display coordinates, and headings, the position of the section can be changed even if the section layout changes. A method has been proposed for estimating.

In addition, the technique for detecting the similarity between structured documents proposed in Patent Document 2 is applied to sections, and the document structures of sections before and after the change are compared, and sections having a more similar structure are compared. A method of mapping each other is conceivable. In particular, since the document structure of the structured document can be expressed in a tree form, the similarity of the structured document is determined using the editing distance of the tree.
JP 2007-293543 A JP2007-052556A

  The first problem is that, when sections of similar layouts are mapped using layout features such as coordinates as in Patent Document 1, a structured document is rendered at least once to obtain layout information. That means you have to get it.

  The second problem is that when the similarity of the document structure is determined and the sections having a more similar structure are mapped as in Patent Document 2 without rendering, the layout is not considered. It means that mapping is not possible.

  An object of the present invention is to provide a structured document processing system, a structured document processing method, and a structured document processing program that can perform section mapping without rendering a structured document.

  It is another object of the present invention to provide a structured document processing system, a structured document processing method, and a structured document processing program that can perform section mapping in consideration of both the layout and structure of a structured document.

  The present invention for solving the above problems is a structured document processing apparatus that calculates dissimilarities between a plurality of structured documents, and each tag constituting the structured document has an influence on the layout. Layout determination means for assigning weights based on the above, and dissimilarity calculation means for calculating dissimilarities between structured documents based on the assigned weights and the tag structure in each structured document. Features.

  The present invention for solving the above-described problems is a structured document processing system that associates a structured document based on the layout and structure of the structured document, and each tag constituting the structured document has an influence on the layout. Layout determination means for assigning weights based on the above, and dissimilarity calculation means for calculating dissimilarities between sections obtained by segmenting the structured document based on the assigned weights and the tag structure of the structured document. The mapping means for associating structured documents with each other based on the calculated dissimilarity and generating display information indicating the association between the structured documents, and the generated display information via the communication network And an information distribution means for transmitting to the terminal.

  The present invention for solving the above problem is a structured document processing method for calculating dissimilarity between a plurality of structured documents, and each tag constituting the structured document has an influence on the layout. A layout determination step for assigning weights based on: and a dissimilarity calculation step for calculating dissimilarities between structured documents based on the assigned weights and the tag structure in each structured document. Features.

  The present invention for solving the above problems is a program for calculating the dissimilarity between a plurality of structured documents, and the program is provided in each tag constituting the structured document in the information processing apparatus. Layout determination processing for assigning weights based on the degree of influence on layout, and dissimilarity calculation processing for calculating dissimilarities between structured documents based on the assigned weights and the tag structure in each structured document Are executed.

  According to the present invention, section mapping can be performed without rendering. The reason is that not the layout but the section configuration tag is to be compared.

  Further, according to the present invention, mapping based on both the layout and structure of the section can be performed. The reason is that the degree of influence of each tag on the layout and the tag structure of the section are used as determination elements.

  In order to explain the features of the present invention, it will be specifically described below with reference to the drawings.

  A feature of the structured document processing system according to the present invention is that it includes a layout determination unit 102, a dissimilarity calculation unit 103, and a mapping unit 104.

  The layout determination unit 102 assigns weights to tags based on the degree of influence of each tag of the structured document on the layout.

  The dissimilarity calculation unit 103 calculates the dissimilarity between sections to be compared based on the tag structure and the weight assigned to each tag by the layout determination unit 102.

  The mapping unit 104 appropriately maps the structured document based on the dissimilarity calculated by the dissimilarity calculating unit 103.

  FIG. 1 is a block diagram showing an example of the structure of a structured document processing system according to the present invention. In the present embodiment, the structured document processing system can be configured by hardware. However, in the following description, the structured document processing system will be described using a case where the structured document processing system is realized by an information processing terminal such as a personal computer that operates according to a program. The structured document processing system may be applied to a business model such as a system that distributes a structured document by dividing it into a plurality of sections. In this case, the structured document processing system may include, for example, a user terminal such as a mobile phone, a PDA, or a personal computer equipped with software for rendering the structured document, and a structured document processing server that processes the structured document. Good.

  As shown in FIG. 1, in the present embodiment, the structured document processing system includes a data processing device 10 that operates under program control, and a storage device 11 that stores information.

  Specifically, the data processing apparatus 10 is realized by a personal computer, a server, or the like that operates according to a program.

  The data processing apparatus 10 includes a document input unit 100, a document analysis unit 101, a layout determination unit 102, a dissimilarity calculation unit 103, a mapping unit 104, and an output unit 105. The storage device 11 is specifically realized by a memory, a hard disk device, or the like. The storage device 11 includes a section storage unit 110 and a mapping storage unit 111.

  The document input unit 100 has a function of acquiring a structured document from the outside and outputting it to the document analysis unit 101. For example, the document input unit 100 reads a structured document from the storage device 11 and outputs it to the document analysis unit 101 in accordance with a user operation. For example, the document input unit 100 receives a structured document (for example, Web content) via a communication network such as the Internet, and outputs the structured document to the document analysis unit 101.

  The document analysis unit 101 has a function of analyzing a structured document acquired from the document input unit 100, extracting a plurality of sections, and storing them in the section storage unit 110. The acquired structured document may be stored in the section storage unit 110 as a single section as it is.

  The layout determination unit 102 has a function of acquiring a section from the section storage unit 110 and assigning a weight to each tag based on the degree of influence of the tags constituting each section on the layout. The layout determination unit 102 has a function of outputting a section in which a weight is assigned to a tag to the dissimilarity calculation unit. The section described here may include a section group (including the structured document itself) that is a set of sections.

  For example, the layout determination unit 102 acquires a tag layout definition from a DTD (Document Type Definition), and determines the degree of influence of each tag on the layout of block elements (basic elements such as headings and paragraphs) and inline elements. (Elements that give roles and functions to display information such as emphasis and links) are classified into two types. Then, the layout determination unit 102 performs weighting based on the degree of influence on the layout by giving a large weight to the block element while giving a small weight to the inline element.

  The dissimilarity calculation unit 103 acquires sections to be compared (at least two sections) from the layout determination unit 102, and calculates dissimilarity indicating the similarity of the structure between sections based on the tag structure of the sections. It has a function to do. Further, the dissimilarity calculation unit 103 has a function of outputting the dissimilarity between sections to the mapping unit 104.

  For example, the dissimilarity calculation unit 103 converts the tag structure of each section into a tree type, and calculates the edit distance between the trees. At this time, the dissimilarity based on both the influence on the layout and the section structure is calculated by using the weight calculated by the layout determination unit 102 as the editing cost of each node of the tree.

  The mapping unit 104 has a function of acquiring the dissimilarity of each section from the dissimilarity calculating unit 103 and mapping (associating) the combination of sections having the smallest dissimilarity, that is, the most similar. Further, the mapping unit 104 has a function of storing the section mapping result in the mapping storage unit 111. For example, in the comparison between the structured document Da and the structured document Db, the mapping unit 104 maps all the sections of Da and Db in order from the combination of sections having the smallest dissimilarity. Further, a section that cannot be mapped due to a difference in the number of sections is mapped as an empty section φ. Note that the mapping unit 104 may divide the set of sections in the structured document into two groups based on the mapped sections in consideration of the relative positions of the sections, and perform mapping within each group. The mapping unit 104 may consider the absolute position of the section, divide the set of sections into a plurality of groups based on the hierarchical structure of the sections in the structured document, and perform mapping within each group.

  The output unit 105 has a function of outputting the section mapping result stored in the mapping storage unit 111 to the outside as display information. For example, the output unit 105 extracts a section designated by the user and a section mapped with the section from the mapping storage unit 111 and displays the extracted section on a display device such as a liquid crystal display unit or a display device. Or sent to the information terminal via the network.

  The storage device 11 includes a section storage unit 110 and a mapping storage unit 111.

  The section storage unit 110 stores a section of the structured document analyzed by the document analysis unit 101 for each structured document.

  The mapping storage unit 111 stores the section mapping result calculated by the mapping unit 104.

  Next, the operation will be described. FIG. 2 is a flowchart illustrating an example of a process in which the structured document processing system maps sections based on the layout and the document structure.

  First, the document input unit 100 acquires a structured document. For example, the document input unit 100 reads a structured document stored in the storage device 11. For example, the document input unit 100 receives a structured document (for example, Web content) via a communication network.

  Next, the document analysis unit 101 analyzes the structured document acquired from the document input unit 100, extracts a plurality of sections, and stores them in the section storage unit 110 for each structured document (step S11).

  Subsequently, the layout determination unit 102 acquires a section from the section storage unit 110, and assigns a weight to each tag based on the degree of influence of the tag on the layout (step S12).

  Next, the dissimilarity calculation unit 103 acquires a section of each structured document from the layout determination unit 102, and based on the weight assigned to the tag and the tag structure, the comparison source section and other sections Each dissimilarity between is calculated (step S13).

  Subsequently, the mapping unit 104 acquires the dissimilarity of the section from the dissimilarity calculation unit 103, maps the sections of the plurality of structured documents based on the dissimilarity, and stores them in the mapping storage unit 111. (Step S14).

  The output unit 105 outputs the section mapping result stored in the mapping storage unit 111.

  As described above, according to the present embodiment, a weight is assigned to a tag based on the degree of influence on the layout, the dissimilarity is calculated based on the weight and the tag structure of the section, and based on the dissimilarity By mapping sections, appropriate section mapping based on layout and document structure can be performed without rendering.

  For example, in the comparison of two structured documents, if there is a section with a slightly different layout, if the type and structure of the tags that make up a section are the most similar compared to other sections, Can be mapped. Further, the similarity that can only be grasped qualitatively can be quantitatively grasped in the form of dissimilarity.

  Next, a specific embodiment of the structured document processing system according to the present invention will be described.

  First, a first embodiment of the structured document processing system will be described. Note that the structured document processing system in this embodiment corresponds to the structured document processing system shown in the first embodiment. In this embodiment, it is assumed that the data processing device is a personal computer and the data storage device is a magnetic disk device.

  The personal computer (data processing apparatus) includes a central processing unit that functions as a document input unit, a document analysis unit, a layout determination unit, a dissimilarity calculation unit, a mapping unit, and an output unit. The magnetic disk device (storage device) stores section information and dissimilarity information analyzed or calculated by a personal computer. The data processing apparatus may be a server, a mobile phone, or the like, and does not depend on the type of terminal. In this embodiment, Web content is targeted as an example of a structured document. For example, a personal computer receives Web content via the Internet.

  In this embodiment, first, the central processing unit receives Web content, analyzes the received Web content, and extracts a section. The central processing unit stores the extracted section information in the magnetic disk device. The received Web content may be used as a single section as it is. In addition, as long as the structured document is created for rendering, it does not depend on the type such as HTML or XML. In this embodiment, HTML is handled as a structured document.

  FIG. 3 is a diagram illustrating an example of a document structure of one structured document, and FIG. 4 is a diagram illustrating a layout configuration of sections after two structured documents are analyzed and a plurality of sections are extracted. . In the present embodiment, the central processing unit extracts a plurality of sections from the structured document D1 as shown in FIG. 3, and numbers such as sections 1 to 8 based on the section extraction order as shown in FIG. Is allocated. A section is similarly extracted from the other structured document D2, and the section information is stored in the magnetic disk device.

  Next, the central processing unit acquires section information from the magnetic disk device, and assigns weights to the tags of each section based on the degree of influence on the layout. In this embodiment, the central processing unit assigns weights to each tag to display information such as block elements (basic elements constituting a layout such as headings and paragraphs) and inline elements (emphasis and links) based on DTD. It is divided into two types of elements and elements that give functions). The section acquired from the magnetic disk device is not limited to the section obtained by analyzing the structured document, but may be another section or the structured document itself. Further, a plurality of sections or structured documents may be acquired.

  FIG. 5 is an explanatory diagram showing an example in which the tags of the structured document D1 in FIG. 3 are divided into block elements and inline elements, and 100 and 1 are assigned as weights, respectively. The weight value in FIG. 5 is “100” for “div” indicating a block element, “h”, “a”, “img” indicating an inline element because the influence on the layout is strongly evaluated. For example, if the influence on the layout is evaluated loosely, the block element is set to “3” and the inline element is set to “1”. The weight difference may be reduced. Further, if it is desired to strongly evaluate the similarity of the tag structure rather than the layout, both the block element and the inline element may be set to 1 to eliminate the weight difference. In addition, the types of tags that determine the degree of influence on layout are not defined as block elements or inline elements, but are defined in other DTD definitions or in structured documents other than DTD such as CSS (Cascading Style Sheets). The layout definition that defines the display of elements may be followed. Moreover, what the user defined in advance may be used. Further, the degree of influence of each tag on the layout when rendered in advance may be calculated and based on the stored result.

  Next, the central processing unit calculates an edit distance from the section to be compared based on the tag structure of the section. In this embodiment, the tag structure is converted into a tree type, and the edit distance of the tree structure is calculated. Furthermore, the dissimilarity is calculated from the edit distance based on the edit cost of one node as the weight of each tag.

  FIG. 6 is an explanatory diagram showing an example in which section 1 of FIG. 5 is converted into a tree type. Note that R in FIG. 6 indicates the root of the tree. In FIG. 6, a node having a large weight value has a large display size, and a node having a small weight value is small.

  FIG. 7 is an explanatory diagram illustrating a calculation example of the edit distance and the dissimilarity with each of the sections 2, 3, and 4, which are different sections, using the tree of section 1 in FIG. 6 as a comparison source. In this embodiment, the calculation distance of the tree is calculated based on the influence on the layout, so if the layout elements (block elements, inline elements, etc.) are the same, they are not related to the tag name. In this embodiment, substitution is not used but substitution = deletion + insertion. Furthermore, in this embodiment, it is considered that the sibling order of the tree also affects the layout, and structural matching by exchanging sibling elements of the tree is not permitted.

  For example, in the calculation of the edit distance with the tree of (a) showing the tree structure of section 2, it is equivalent to the structure in which two inline elements “a” and “img” are deleted (del) from the comparison source tree Therefore, the edit distance is 2, and the dissimilarity based on the weight is 2.

  Similarly, the tree in (b) showing the tree structure of section 3 is equivalent to the structure in which one inline element is deleted, one is inserted (ins), and one block element is deleted from the comparison source tree. Therefore, the edit distance is 3, and the dissimilarity based on the weight is 102.

  Furthermore, since the tree in (c) showing the tree structure of section 4 is equivalent to the structure in which two inline elements are deleted and inserted from the comparison source tree, the edit distance is 4, which is based on the weight. The dissimilarity is 4.

  In order to make the similarity determination strict, replacement may be used by giving a certain weight to the editing cost if the tag name is different even for the same element. In addition, in order to ease the similarity determination, structure matching may be permitted by replacing the sibling elements of the tree. If the tag structure is used, it is not necessary to convert it to a tree type. For example, assuming that the block element is 1 and the inline element is 0, the tag structure may be converted into a binary expression, and the edit distance of the converted character string may be obtained. Further, the central processing unit may finish the process when the dissimilarity is calculated. In this case, the information stored in the magnetic disk device is dissimilarity.

  Next, the central processing unit maps in order from the combination of sections having the smallest dissimilarity. In the present embodiment, the process of dividing the section into two groups based on the mapped section and mapping the sections in each group is repeated.

  FIG. 8 is an explanatory diagram showing an example of mapping between the structured documents D1 and D2 of FIG. For example, first, a combination of section 4 (D1) and section 4 (D2) having the smallest dissimilarity is mapped. This mapping divides the structured document into two groups, G11 and G12 for D1, and G21 and G22 for D2.

  Next, a combination of sections having the smallest dissimilarity in each group is mapped. For example, as shown in FIG. 9, in G11: G21, section 2 (D1) and section 3 (D2), and in G12: G22, combinations of section 8 (D1) and section 9 (D2) are mapped.

  In D1, the group G11 is divided into G111 and G112, G12 is divided into G121 and G122, while in D2, G21 is divided into G211 and G212, and G22 is divided into G221 and G222. G122, G212, and G222 are empty groups in which no section exists.

  Subsequently, a combination of sections having the smallest dissimilarity in each group is mapped as before. For example, as shown in FIG. 10, in G111: G211, a combination of section 1 (D1) and section 1 (D2), and in G121: G221, a combination of section 5 (D1) and section 7 (D2) is mapped.

  In D1, the group G111 is divided into G1111 and G1112, G121 is divided into G1211 and G1212, while in D2, G211 is divided into G2111 and G2112, and G221 is divided into G2211 and G2212. Here, in G112: G212, since the partner of the combination of section 3 (D1) does not exist in D2, it is mapped to the section φ that does not exist.

  Similarly, the mapping in the divided groups is repeated, and finally the mapping result as shown in FIG. 11 is obtained.

  In addition, as shown in FIG. 12, group division may be performed using a hierarchical structure in a section document.

  For example, when searching to a depth of 2, first, group mapping is performed in a hierarchy of a depth of 1, as shown in FIG. Next, as shown in FIG. 13B, group mapping is performed in a hierarchy of depth 2. After that, section mapping is performed in each divided group.

  In the above-described invention, in an application that uses a structured document having a layout by dividing it into a plurality of sections, it can be appropriately mapped to similar sections based on the layout and the document structure. Mapping that is resistant to change is possible. In addition, the similarity of structured documents can be grasped quantitatively.

  For example, in a web content such as a blog where the layout configuration changes frequently, an application that manages the update information of a specific section deletes or adds other sections, and even if the section configuration changes, the specific section is Can be managed consistently.

  Further, even an application that manages update information of the entire Web content can identify which section has been deleted or added. Furthermore, since the similarity can be grasped in section units even between Web contents having completely different URLs, it is possible to perform processing such as hiding a section having a specific similarity in all the Web contents.

It is a block diagram which shows the other structural example of a structured document processing system. It is a flowchart which shows an example of the process in which a structured document processing system maps a section. It is explanatory drawing which shows the example of the document structure of a structured document. It is explanatory drawing which shows the layout structure of a section. It is explanatory drawing which shows the example which allocated 100 and 1 as a weight with respect to the tag of a block element and an inline element, respectively. It is explanatory drawing which shows the example which converts a section into a tree type. It is explanatory drawing which shows the example of calculation of edit distance and dissimilarity. It is explanatory drawing which shows the mapping procedure of structured document D1 and D2. It is explanatory drawing which shows the mapping procedure of structured document D1 and D2. It is explanatory drawing which shows the mapping procedure of structured document D1 and D2. It is explanatory drawing which shows the example which mapping of the structured documents D1 and D2 was completed. It is explanatory drawing which shows the example of a hierarchical structure of the structured document D1. It is explanatory drawing which shows the example of mapping using the hierarchical structure of the structured documents D1 and D2.

Explanation of symbols

10 data processing device 11 storage device 100 document input unit 101 document analysis unit 102 layout determination unit 103 dissimilarity calculation unit 104 mapping unit 105 output unit 110 section storage unit 111 mapping storage unit

Claims (39)

A structured document processing apparatus that calculates dissimilarity between a plurality of structured documents,
Layout determination means for assigning a weight to each tag constituting the structured document based on the degree of influence on the layout;
A structured document processing apparatus comprising: a dissimilarity calculating unit that calculates dissimilarity between structured documents based on the assigned weight and the tag structure in each structured document.   2. The structured document processing apparatus according to claim 1, further comprising mapping means for associating structured documents having a small calculated dissimilarity.   3. The structured document processing apparatus according to claim 2, wherein the mapping means associates a plurality of structured documents in order from structured documents having a low dissimilarity.   3. The mapping unit according to claim 2, wherein the mapping unit divides each structured document into a plurality of groups on the basis of the first associated structured document, and associates the structured documents in each group with each other. 3. The structured document processing apparatus according to 3.   The dissimilarity calculating means calculates dissimilarity between sections of each structured document based on the assigned weight and a tag structure in a section obtained by subdividing the structured document. The structured document processing apparatus according to claim 1.   6. The structured document processing apparatus according to claim 5, further comprising document analysis means for analyzing the input structured document and extracting a section.   7. The structured document processing apparatus according to claim 1, wherein the layout determination unit uses a layout definition predetermined for the structured document as a weight calculation criterion.   7. The structured document processing apparatus according to claim 1, wherein the layout determination unit uses a definition predetermined by the user as a weight calculation criterion.   The layout determination means measures the influence of each tag on the layout when pre-rendering as a weight calculation criterion, and uses a definition defined by the system based on the measurement. The structured document processing apparatus according to any one of the above.   10. The structured document processing apparatus according to claim 1, wherein the dissimilarity calculating unit uses a structure obtained by converting a tag structure into a tree type for calculating the dissimilarity.   10. The structured information according to claim 1, wherein the dissimilarity calculating means uses a structure obtained by converting a tag structure into a single-byte character string for calculating the dissimilarity. Document processing device.   12. The structured document processing apparatus according to claim 10, wherein the dissimilarity calculating means uses the edit distance of the structure for calculating the dissimilarity. A structured document processing system that associates structured documents based on the layout and structure of a structured document,
Layout determination means for assigning a weight to each tag constituting the structured document based on the degree of influence on the layout;
Dissimilarity calculation means for calculating dissimilarity between sections obtained by subdividing the structured document based on the assigned weight and the tag structure of the structured document;
Mapping means for associating structured documents with each other based on the calculated dissimilarity, and generating display information indicating the association between the structured documents, and the generated display information via an information terminal A structured document processing system comprising: an information distribution means for transmitting to the network. A structured document processing method for calculating dissimilarity between a plurality of structured documents,
A layout determination step for assigning a weight to each tag constituting the structured document based on the degree of influence on the layout;
A structured document processing method comprising: a dissimilarity calculation step of calculating dissimilarity between structured documents based on the assigned weight and the tag structure in each structured document.   15. The structured document processing method according to claim 14, further comprising a mapping step of associating structured documents having a small calculated dissimilarity.   16. The structured document processing method according to claim 15, wherein the mapping step associates a plurality of structured documents in order from structured documents having a low dissimilarity.   16. The mapping step according to claim 15, wherein the mapping step divides each structured document into a plurality of groups on the basis of the first associated structured document, and associates the structured documents in each group with each other. The structured document processing method according to claim 16.   The dissimilarity calculating step calculates a dissimilarity between sections of each structured document based on the assigned weight and a tag structure in a section obtained by subdividing the structured document. The structured document processing method according to claim 14.   19. The structured document processing method according to claim 18, further comprising a document analysis step of extracting a section by analyzing the input structured document.   10. The structured document processing method according to claim 14, wherein the layout determination step uses a layout definition predetermined for the structured document as a weight calculation criterion.   20. The structured document processing method according to claim 14, wherein the layout determination step uses a definition predetermined by the user as a weight calculation criterion.   The layout determination step measures the influence of each tag on the layout when pre-rendering as a weight calculation criterion, and uses a definition defined by the system based on the measurement. The structured document processing method according to any one of the above.   The structured document processing method according to any one of claims 14 to 22, wherein the dissimilarity calculation step uses a structure obtained by converting a tag structure into a tree type for calculating the dissimilarity.   23. The structured document according to claim 14, wherein the dissimilarity calculating step uses a structure in which a tag structure is converted into a single-byte character string for calculating the dissimilarity. Processing method.   25. The structured document processing method according to claim 23, wherein the dissimilarity calculation step uses a structure edit distance for calculating the dissimilarity.   The structured document processing method according to any one of claims 15 to 25, further comprising an information transmission step of transmitting a result of associating structured documents with each other through a communication network in the mapping step. A program for calculating the dissimilarity between a plurality of structured documents, the program in an information processing device,
Layout determination processing for assigning a weight to each tag constituting a structured document based on the degree of influence on the layout;
A non-similarity calculation process for calculating a non-similarity between structured documents based on the assigned weight and the tag structure in each structured document.   28. The program according to claim 27, wherein a mapping process for associating structured documents with a low calculated dissimilarity is executed.   29. The program according to claim 28, wherein the mapping process associates a plurality of structured documents in order from structured documents having a low dissimilarity.   30. The mapping process according to claim 28, wherein the mapping process divides each structured document into a plurality of groups on the basis of the structured document associated first, and associates the structured documents in each group with each other. 29. The program according to 29.   The dissimilarity calculation processing calculates a dissimilarity between sections of each structured document based on the assigned weight and a tag structure in a section obtained by subdividing the structured document. The program according to any one of claims 27 to 30.   32. The program according to claim 31, wherein a document analysis process for analyzing the input structured document and extracting a section is executed.   The program according to any one of claims 27 to 32, wherein the layout determination processing uses a layout definition predetermined for the structured document as a weight calculation criterion.   The program according to any one of claims 27 to 32, wherein the layout determination process uses a definition predetermined by a user as a weight calculation criterion.   The layout determination process uses a definition defined by the system based on measurement of an influence degree of each tag on the layout when rendering in advance as a weight calculation criterion. A program according to any of the above.   36. The program according to claim 27, wherein the dissimilarity calculation process uses a structure obtained by converting the tag structure into a tree type for calculating the dissimilarity.   The program according to any one of claims 27 to 35, wherein the dissimilarity calculation processing uses a structure obtained by converting a tag structure into a single-byte character string for calculation of dissimilarity.   38. The program according to claim 36 or claim 37, wherein the dissimilarity calculation processing uses the edit distance of the structure for calculating the dissimilarity.   The program according to any one of claims 28 to 38, wherein an information transmission process for transmitting a result of association in the mapping process via a communication network is executed.

Images