JP2012178079A - Document evaluation support system and document evaluation support method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a document evaluation support system for visualizing a place where the content of an input document is related with standard document structured data of preliminarily defined arbitrary document content.SOLUTION: This document evaluation support system includes: a standard document structured data storage device for storing standard document structured data about the content of an arbitrary document; a structured document conversion device for converting an input document into document structured data; a document structured data storage device for storing the document structured data converted by the structured document conversion device; a structured data comparison and evaluation device for comparing and evaluating the standard document structured data and the input document structured data; a reference information creation device for extracting an attention required place from the input document, and for creating reference information corresponding to the attention required place; an evaluation result storage device for calculating and storing the evaluation scores of respective words and phrases from the comparison result of the standard document structured data and the input document structured data; and an evaluation result display device for changing the display of a graph or words and phrases in accordance with the evaluation result.

Description

  In the present invention, the relationship between words in an input document is similar to the above in order to indicate where the content described in the input document is a related location in a standard knowledge network structure related to the content of the document. This is related to a document evaluation support system that supports similarity evaluation by matching with a standard knowledge network structure and displaying related parts according to the evaluation result.

  Traditionally, sentences in a document are described as structured data, information is searched from other databases, and the search results are judged by the user based on experience and discretion, and the results are used. .

  An ontology is known as a technique for systematically expressing a concept on a certain knowledge base such as a phrase about an arbitrary content and a relation between phrases. In addition, as a format for describing structured data representing an ontology, there are an XML (eXtensible Makeup Language) format and an RDF (Resource Description Framework).

  Information retrieval systems and structured data retrieval programs such as those disclosed in [Patent Document 1] and [Patent Document 2] are known as methods for retrieving information using an inquiry statement described in an ontology description format.

  There is an information search system as shown in [Patent Document 3] as a method of searching a document by a search request using an ontology and feeding back evaluation information about a search result.

JP 2005-165958 A JP 2010-79857 A JP 2003-108597 A

  In a conventional search method, a search phrase or search sentence and information related to the phrase or sentence are searched from other documents or databases.

  However, how much is the description content of the input document related to which part in the standard knowledge network structure related to the content of the document or the knowledge network structure representing the content that the document should be described as standard? In order to know which part is described, which part is not described, etc., not only the words and phrases between the contents of the input document but also the relation between the words and phrases are searched. The knowledge network structure of the input document and the knowledge network structure of the input document, search for related parts, evaluate the similarity, and change the display of each word constituting the standard knowledge network structure according to the evaluation result is required.

  In order to solve the above-mentioned problem, in the present invention, a word and phrase related to the contents of an arbitrary document, means for storing a standard knowledge network structure representing the relationship between two phrases, and an input document as described above, Means for expressing a relationship between words and phrases, and two phrases, and converting to a knowledge network structure composed of the words included in the document; a knowledge network structure of a sentence selected from an input document; A means for comparing the knowledge network structure and searching for a match or similar part, a means for evaluating the similarity as a score of each word, and a display of each word constituting the standard network structure in the above evaluation result It is characterized by comprising means for changing in response.

  Further, in the document evaluation support system of the present invention, the means for converting the input document into a knowledge network structure parses each sentence of the input document to show the relationship between the subject, predicate, object, etc. And the relationship thereof, and further converted into a relationship similar to a standard knowledge network structure.

  In the document evaluation support system of the present invention, the location where the knowledge network structure of the input document matches the standard knowledge network structure is the same as the knowledge network structure of each sentence in the input document. The knowledge network structure of the input document is similar to the knowledge network structure of the input document because the knowledge network structure of the input document is similar to the knowledge network structure of the input document. When a knowledge network structure interpolating other words and relations exists in the standard knowledge network structure in a state where the relation between words and phrases can be maintained between words, or in each input document A knowledge network structure that has been deleted in a state that maintains the relationship between phrases in the knowledge network structure of the sentence It is characterized in that it is when present in the network structure.

  According to the document evaluation support system of the present invention, it is possible to grasp the related parts and features of the description content of the input document in the standard knowledge network structure about the content of an arbitrary document.

Basic configuration diagram. Processing procedure for extracting, evaluating, and displaying matching and similar parts of input document structured data from standard document structured data. Standard document structured data example. An example in which a point requiring attention is incorporated into standard document structured data. Processing procedure for extracting important points of input document structured data from standard document structured data and outputting reference information. Example of converting input document to input document structured data. An example of the process sequence in a structured data comparison and evaluation apparatus. The example of the location in the standard document structured data that matches the structured data of an arbitrary sentence in the input document. The example of the location similar to the structured data of the arbitrary sentences of an input document in standard document structured data. The example of the location similar to the structured data of the arbitrary sentences of an input document in standard document structured data. An example of evaluating a portion in the standard document structured data that matches the structured data of an arbitrary sentence in the input document. An example of evaluating a portion similar to structured data of an arbitrary sentence in an input document in standard document structured data. An example of evaluating a portion similar to structured data of an arbitrary sentence in an input document in standard document structured data. The example of a screen which displays the evaluation result about the selection range of an input document in standard document structured data.

  An embodiment of an information reference support system according to the present invention will be described below with reference to FIGS.

  FIG. 1 shows the basic configuration of the document evaluation support system of this embodiment. The system includes a standard document structure database (101) for storing standard network structure data (hereinafter referred to as standard document structured data) for arbitrary document contents, and an input document (103) as knowledge network structure data ( (Hereinafter referred to as input document structured data), a structured document conversion device (104) for conversion to an input document structured data, an input document structured database (105) for storing input document structured data converted by the structured document conversion device, Structured data comparison / evaluation to extract points of caution by comparing or evaluating the standard document structured data and the input document structured data. Device (106), evaluation result database (accumulation result for each sentence of input document structured data) ( 09), a reference information database (107) for storing reference information about a point requiring attention, a reference information creation device (108) for creating corresponding reference information when a point of interest is extracted from the input document, It is configured by an evaluation result display device (110) that changes the display of graphs and phrases according to the evaluation results and displays them on a screen or the like, or displays reference information.

  FIG. 2 shows a processing procedure for extracting, evaluating, and displaying matching and similar parts of the input document structured data from the standard document structured data. An example of a processing procedure in the document evaluation support system of the present invention shown in FIG. 1 is shown in a flowchart.

  After the start (step 201), the standard document structured database is read (step 202). The standard document structured database is defined in advance according to the content of the target document. For example, in a contract document such as a technical specification for a certain product, the required terms and terms for components and implementation items, etc. from the viewpoint of what should be described as the technical specification for that product. A predefined mutual relationship between words (this is called a relationship) is described in a knowledge network structure.

  Next, the input document is read (step 203), and a target range to be evaluated this time is designated from the input document (step 204). The document is composed of one or more sentences, and the designated target range is also composed of one or more sentences. The sentence in the designated target range is converted into a knowledge network structure similar to the standard document structured data, and input document structured data is created (step 205). The conversion method will be described later.

  Next, in order to evaluate the coincidence and similarity of the input document structured data with the related parts in the standard document structured data, knowledge network structure data of each sentence for all sentences in the input document structured data Are retrieved from the standard document structured data (step 206) and evaluated based on the retrieval status (step 207). The search method will be described later with reference to FIG. 8, and the evaluation method (matching and similar evaluation) will be described later with reference to FIGS. At this time, the evaluation score of the word / phrase in the standard document structured data is increased for the word / phrase having a high degree of similarity due to the evaluation of each sentence. When the evaluation is completed for all sentences in the input document structured data (step 208), the evaluation scores of the words and phrases in the standard document structured data are summed up, and the result is displayed (step 209), and the process is completed (step 209). Step 210).

  FIG. 3 shows an example of the standard document structured data 101 shown in FIG. 1 in the document evaluation support system of this embodiment.

  In this example 300, for the contents to be described in the technical specification of the product “T100”, the words and phrases and the relationship thereof are represented as tree-like data. Enclosed in an ellipse are the words that are normally included in the contents of the technical specification, and the relationship between the words is connected with a line, and a definition word representing the relationship is written near the line. Definition terms that conceptually represent relationships are defined in advance. In this example, p / o (part_of) indicates the device configuration, and a / o (attribute_of) indicates the attribute, type, and action item. . Thus, for example, one of the components in the technical specification “T100” 301 is “Compressor” 302. Here, the standard document structured data is represented as tree-like data. However, as long as the relationship between words and phrases and the relationship between the two words can be defined, for example, XML, RDF, or table format may be used.

  FIG. 4 shows an example in which the information of the caution area database 102 is specified in the standard document structured data 101 shown in FIG. 3 in the document evaluation support system of this embodiment.

  In this example 400, if a content such that the attribute of “Fuel” 401 is “Gas” 402 is described in the technical specification, it indicates a point requiring attention (403), and reference information (404) is displayed. Linked. For example, a format in which reference information No. is added to data indicating the phrase “Fuel”, the phrase “Gas”, and the relationship (a / o) between the two phrases is conceivable. At this time, when the phrase “Fuel”, the phrase “Gas”, and the relationship (a / o) between the two phrases are extracted from the input document structured data, the linked reference stored in the reference information database. A method of utilizing information can be considered. When extracting a part that matches or is similar to the input document structured data from the standard document structured data, it is possible to search for a part requiring attention.

  FIG. 5 is a flowchart illustrating an example of a processing procedure in the document evaluation support system according to the present exemplary embodiment when the point of interest 102 is clearly specified in the standard document structured data 101 illustrated in FIG. 4.

  In FIG. 5, determination of a point requiring attention (step 211) and processing for outputting reference information when a point requiring attention (step 212) are added to the flowchart shown in FIG.

  FIG. 6 shows an example of a structured document conversion method executed by the standardized document conversion apparatus 104 shown in FIG. 1 in the document evaluation support system of this embodiment.

  For example, as a result of performing a syntax analysis used in a formal language analysis such as natural language processing on a sentence as shown in Sentence Example 1 (601) described in an input document, a grammatical phrase dependency relationship I understand. In this example, it is assumed that the subject “panel” (602), the predicate “be provided” (603), and the object “device” (604) are extracted as a result of the syntax analysis (605). Furthermore, using the verb-relationship conversion table (606), the predicate (verb) is defined based on the standard document structure data according to the content of the target document, and the phrase “panel”, the phrase “device”, and The relationship between the two phrases is converted to “p / o” (607). Similarly, in the sentence example 2 (611), when the subject “heater” (612), the predicate “have” (613), the object “KOSHA” (614), and “ASME” (615) are extracted as the result of the syntax analysis. (616), using the verb-relative conversion table (606), the phrase “heater” and the phrase “KOSHA”, and the relationship between the two phrases “a / o”, the phrase “heater”, the phrase “KOSHA”, and 2 It is converted into the relationship “a / o” of two words (617). As a result, the input document is in a format that can be easily compared and evaluated with the standard document structured data.

  FIG. 7 is a flowchart showing an example of a processing procedure in the structured data comparison / evaluation apparatus 106 shown in FIG. 1 in the document evaluation support system of the present invention.

  After the start (step 701), the input document structured data is read (step 702). Further, the structured data of the target sentence is read from among them (step 703). Structured data of a sentence, for example, the phrase “panel” and the phrase “device” converted from the sentence example 1 and the relationship “p / o” (607) between the two phrases, and the phrase “heater” converted from the sentence example 2 ”And the phrase“ KOSHA ”, and the relationship“ a / o ”between the two phrases, the phrase“ heater ”, the phrase“ KOSHA ”, and the relationship“ a / o ”(617) between the two phrases. Next, all parent-child relationships are extracted from the structured data of the sentence (704). For example, the sentence example 1 has one parent-child relationship, but the sentence example 2 has two parent-child relationships. The parent-child relationship includes a parent word and a child word and a relationship between two words. Further, each parent-child relationship is extracted from the standard structured data (705, 706). This is repeated for the parent-child relationships extracted from all sentences in the target document (707, 708, 709), and the process ends.

  FIG. 8 shows the input document structured data from the standard document structured data, which is executed by the structured data comparison / evaluation apparatus 106 shown in FIG. 1 according to the processing procedure shown in FIG. 7 in the document evaluation support system of the present invention. An example of searching is shown.

  The sentence example 1 (601) illustrated in FIG. 6 is converted into the phrase “panel”, the phrase “device”, and the relationship “p / o” between the two phrases (607, 810). The word / phrase and the relationship between the two words / phrases that match this are searched from the standard document structured data. As a result, in this example, in the standard document structured data, a phrase and a phrase that match the sentence example and a parent-child relationship indicated by the relationship between the two phrases are searched (821).

  FIG. 9 shows the input document structured data from the standard document structured data, which is executed by the structured data comparison / evaluation apparatus 106 shown in FIG. 1 according to the processing procedure shown in FIG. Another example of searching is shown.

  Sentence example 2 (611) illustrated in FIG. 6 includes the phrase “heater” and the phrase “KOSHA”, the relationship between the two phrases “a / o”, the phrase “heater”, the phrase “ASME”, and the relationship between the two phrases. It is converted to “a / o” (617, 910). The word / phrase and the relationship between the two words / phrases that match this are searched from the standard document structured data. At this time, the phrase “Heater” and the phrase “Certificate” of the standard document structured data and the relationship “a / o” between the two phrases, the phrase “Certificate”, the phrase “KOSHA”, and the relationship “a” between the two phrases / O "(921), since the two relations are the same, the relation between the two phrases for" Heater "and the phrase" KOSHA "is inherited to become" a / o "(931, 911). That is, when one of the attributes (implementation items) of “Heater” is “Certificate” and one of the attributes (types) of “Certificate” is “KOSHA”, one of the attributes (implementation items) of “Heater” is “ KOSHA "(911). As a result, in this example, in the standard document structured data, the phrase and phrase that match the sentence example and the parent-child relationship indicated by the relationship between the two phrases are retrieved (931). The relationship between “Heater” and “ASME” is also the same (912, 922, 932).

  In this way, by deleting or omitting a phrase between two phrases so that the mutual relationship between the two phrases is maintained for any sentence structured data in the input document structured data The phrase in the sentence structured data and the phrase and the case where the relationship between the two phrases having a high correlation with each other coincide with a part of the standard document structured data are regarded as similar parts.

  By displaying the evaluation result of the input document in this way, it is possible to grasp the related parts and features of the input document, and the sufficient and insufficient parts of the description content.

  FIG. 10 shows the input document structured data out of the standard document structured data, which is executed by the structured data comparison and evaluation apparatus 106 shown in FIG. 1 according to the processing procedure shown in FIG. 7 in the document evaluation support system of the present invention. Another example of searching is shown.

  As the structured data of sentence example 3 in the input document structured data, the relationship between the two phrases of the phrase “Fuel” and the phrase “system” is “p / o”, and the phrase “system” and the phrase “Oil” If the relationship between the two phrases is “p / o” (1010), the phrase and the phrase and the relationship between the two phrases that match this are searched from the standard document structured data. At this time, as a pattern of the parent-child relationship to be searched, Example 3 (a) the phrase “Fuel” and the phrase “system”, and the relationship between the two phrases “p / o” (1011), Example 3 (b) the phrase “system” And the phrase “Oil”, the relationship between the two phrases “p / o” (1012), and the two relations are the same “p / o”. Example 3 (c) The phrase “Fuel” and the phrase “ Oil ”and the relationship“ p / o ”(1013) between the two phrases. As a result, in this example, in the standard document structured data, the phrase and the phrase that match Example 3 (c) and the parent-child relationship indicated by the relationship between the two phrases are searched (1021).

  11, 12, and 13, when the structured data comparison and evaluation apparatus 106 illustrated in FIG. 1 retrieves the input document structured data from the standard document structured data in the document evaluation support system of the present invention. The example of calculation of an evaluation score is shown.

FIG. 11 shows an example in which an evaluation score is calculated for sentence example 1 retrieved in FIG.
Parent phrase: 1.0 × α ₀₁ (α ₀₁ : Parent weight when matched)
Child phrase: 1.0 × α ₀₂ (α ₀₂ : Child weight when matched)
As a result, when α ₀₁ = α ₀₂ = 1.0,
Parent phrase “panel”: 1.0 × 1.0 = 1.0
Child phrase “device”: 1.0 × 1.0 = 1.0
FIG. 12 shows an example in which an evaluation score is calculated for sentence example 2 retrieved in FIG.
Parent word / phrase: 1.0 × (1 / n) × β ₁₁ (β ₁₁ : “similar” by word interpolation
Parent weight)
Child phrase: 1.0 × (1 / n) × β ₁₂ (β ₁₂ : “similar” by interpolation of phrases
Child weight)
As a result, when β ₁₁ = β ₁₂ = 1.0, n = 2
Parent phrase “Heater”: 1.0 × 0.5 × 1.0 = 0.5 (1)
Child phrase “KOSHA”: 1.0 × 0.5 × 1.0 = 0.5
Parent phrase “Heater”: 1.0 × 0.5 × 1.0 = 0.5 (2)
Child phrase “ASME”: 1.0 × 0.5 × 1.0 = 0.5
From (1) and (2), the score of “Heater” is 0.5 + 0.5 = 1.0.

Thus, for any sentence structured data in the input document structured data, the sentence structure is added and interpolated so that the mutual relationship between the two phrases is maintained between the two phrases. The case where the relation between the words and phrases in the structured data and the two phrases having high correlation with each other coincides with a part of the standard document structured data is regarded as a similar part. FIG. 13 shows an example in which an evaluation score is calculated for sentence example 3 retrieved in FIG.
Parent word / phrase: 1.0 × (1 / m) × γ ₂₁ (γ ₂₁ : When “similar” by deleting word / phrase
Parent weight)
Child phrases: 1.0 × (1 / m) × γ ₂₂ (γ ₂₂ : “similar” by deleting a phrase
Child weight)
As a result, when γ ₂₁ = γ ₂₂ = 1.0, m = 2
Parent word “Fuel”: 1.0 × 0.5 × 1.0 = 0.5
Child phrase “Oil”: 1.0 × 0.5 × 1.0 = 0.5
In this way, by deleting or omitting a phrase between two phrases so that the mutual relationship between the two phrases is maintained for any sentence structured data in the input document structured data The phrase in the sentence structured data and the phrase and the case where the relationship between the two phrases having a high correlation with each other coincide with a part of the standard document structured data are regarded as similar parts.

  Note that the weights shown here are changed depending on the standard document structured data and the configuration and level of detail of the input document.

  FIG. 14 shows an example of an output screen in which the display of the words constituting the standard structured data is changed according to the evaluation result by the evaluation result display device 110 shown in FIG. 1 in the document evaluation support system of the present invention. In this example, when the target range of the document is selected and the standard document structure is displayed, the evaluation score of the word or phrase related to the content of the target sentence in the entire standard document structured data is increased. Correspondingly, the shades of colors in the ellipses representing words are changed and displayed. As a result, for example, it is easy to know the characteristics of the content of the selected sentence in the contract document, where it relates to the standard contract document, and where the content is insufficient.

  In the first embodiment, it is assumed that the structured document conversion apparatus 104 converts the data. However, if the converted data is recorded in the input document structured database 105, the analysis processing is not necessarily required. That is, the screens of FIGS. 3, 4, and 8 to 14 are displayed by comparing the standard document structured data recorded in the database with the input document structured data by the structured data comparison / evaluation apparatus 106. .

  That is, standard knowledge network structure data (hereinafter referred to as standard document structured data) that describes the relationship between two phrases that are highly related to each other with respect to the contents of a specified type of document. Knowledge network indicating the relationship between words and phrases and two words and phrases having a high relationship with each other, which is stored in the database (101) recorded in advance and has an input arbitrary document in the same format as the standard document structured data Structure data (hereinafter referred to as input document structured data) is stored in a database (105), and a match or similarity between the standard document structured data (101) and the input document structured data (105) is expressed as a phrase. In accordance with the evaluation result, the display of the words constituting the standard document structured data is changed and described in the input document in the standard document structured data. By using the display method of the document evaluation support system that visualizes the relevant parts of the contents being displayed on the display means, the relevant parts and characteristics of the contents of the input document are grasped in the standard knowledge network structure of the contents of any document can do.

  Further, according to the display method of the document evaluation support system in which the evaluation score is displayed in association with the location where the display of the phrase is changed, the operator can combine the related document with the similarity between the standard document structured data and the input document structured data. Can be easily grasped.

  In addition, this invention is not limited to an above-described Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. In addition, a part of the configuration of a certain embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of a certain embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

    Each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Further, each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, files, measurement information, and calculation information for realizing each function is stored in a recording device such as a memory, a hard disk, or an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD. Can be put in. Therefore, each process and each configuration can realize each function as a processing unit, a processing unit, a program module, and the like.

  Further, the control lines and information lines indicate what is considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. Actually, it may be considered that almost all the components are connected to each other.

101 Standard Document Structured Database 102 Important Point Database 103 Input Document 104 Structured Document Conversion Device 105 Input Document Structured Database 106 Structured Data Comparison Evaluation Device 107 Reference Information Database 108 Reference Information Creation Device 109 Evaluation Result Database

Claims (8)

Standard knowledge network structure data (hereinafter referred to as standard document structured data) that describes the relationship between two phrases that are highly related to each other with respect to the contents of a specified type of document. Database and
Knowledge network structure data (hereinafter referred to as input document structured data) indicating the relationship between two phrases that are highly related to each other, in terms of phrases and phrases that are in the same format as the standard document structured data. A structured document conversion device (104) for converting to
The agreement or similarity between the standard document structured data and the input document structured data is evaluated for each word, and the display of the words constituting the standard document structured data is changed according to the evaluation result. A document evaluation support system comprising a structured data comparison / evaluation device for visualizing a related portion of the contents described in an input document in the document structured data. The document evaluation support system according to claim 1,
In the standard document structured data, a point of caution in a specified type of document is described as a relationship between a phrase and a phrase, and two phrases that are highly related to each other,
The structured data comparison / evaluation apparatus searches for a point requiring attention when extracting a part that matches or is similar to the input document structured data from the standard document structured data. . The document evaluation support system according to claim 1,
The input arbitrary document is composed of a plurality of sentences, and each sentence is converted into knowledge network structure data (hereinafter, sentence structured data) indicating the relation between two phrases that are highly related to each other. The input document structured data is composed of one or more sentence structured data,
The structured data comparison / evaluation apparatus selects a sentence to be evaluated when evaluating the match or similarity between the standard document structured data and the input document structured data. Evaluation support system. The document evaluation support system according to claim 1,
The portion that matches the input document structured data in the standard document structured data is a phrase and a phrase of arbitrary sentence structured data in the input document structured data, and two that are highly related to each other. This is a case where the relationship of words / phrases exists in the standard document structured data as well, and the similar part refers to any sentence structured data in the input document structured data (1) between two words / phrases By adding and interpolating words so that the relationship between the two words is maintained, the words and phrases in the sentence structuring data and the relationship between the two words having a high relationship with each other are structured in the standard document. If it matches part of the data, or (2) delete or omit the phrase between the two phrases so that the mutual relationship between the two phrases is preserved. Words and phrases and their interrelationships Document evaluation assistance system, wherein the relationship have two words is if it matches a portion of the standard document structured data. The document evaluation support system according to claim 1,
The structured data comparison / evaluation apparatus determines how much the standard document structured data and the input document structured data match or are similar to each other, a phrase of sentence structured data in the input document structured data, and mutual Evaluation of each phrase that is determined by how the relationship between the two highly related phrases matches the phrase in the standard document structured data and the relationship between the two highly related phrases A document evaluation support system, wherein a score is used to determine the evaluation score of all sentences in the input document structured data to be evaluated for each word. The document evaluation support system according to claim 1,
The structured data comparison / evaluation apparatus relates to an arbitrary document input into the standard document structured data based on a match or similarity result between the standard document structured data and the input document structured data. To display a location, a network diagram representing standard document structured data, a tree-structured phrase, or a display of mutual relations is highlighted by distinguishing it by color, shape, etc. depending on the total evaluation score of each phrase. Document evaluation support system, characterized by visualization. Standard knowledge network structure data (hereinafter referred to as standard document structured data) that describes the relationship between two phrases that are highly related to each other with respect to the contents of a specified type of document. Stored in the database
Knowledge network structure data (hereinafter referred to as input document structured data) indicating the relationship between two phrases that are highly related to each other, in terms of phrases and phrases that are in the same format as the standard document structured data. In the recorded database,
The agreement or similarity between the standard document structured data and the input document structured data is evaluated for each word, and the display of the words constituting the standard document structured data is changed according to the evaluation result. A display method for a document evaluation support system, characterized in that a related portion of contents described in an input document in document structured data is visualized on a display means.   8. The display method for a document evaluation support system according to claim 7, wherein an evaluation score is displayed in association with a place where the display of the phrase is changed.

Images