JP6187745B2 - Document analysis system, method and program - Google Patents

Description

  The present invention relates to a document analysis system, method, and program, and more particularly, to a document analysis system, method, and program for extracting an ambiguous expression from a document written in a natural language.

  2. Description of the Related Art In recent years, document analysis systems for analyzing documents written in a natural language using an information processing apparatus and analyzing preferential correction points and quality of the documents have been developed.

  An example of a technique related to the document analysis system is disclosed in Patent Document 1 as “Japanese sentence correcting device, Japanese sentence correcting method, and Japanese sentence correcting program”. The Japanese sentence correction device disclosed in Patent Document 1 includes a sentence composition database in which rules relating to structures that are easy to machine translate are registered, and a script / term database in which rules relating to letters and terms that are easy to machine translate are registered. And a style database in which rules relating to styles that are easy to machine translate are registered, a correction location extraction means, a display means, a correction means, and an output means.

  The similar expression extraction device having such a configuration operates as follows. That is, the correction part extracting means reads from the original Japanese text to be read, the rules relating to the above configuration, such as “create a sentence by the number of predicates”, and the above-mentioned script characters such as “the sentence does not include a letter, a typo, or a missing letter”. A configuration, script, term, or style that is contrary to the rules relating to the term or the rules relating to the style such as “the sentence does not include an ambiguous phrase” is extracted. Further, the display means displays the extracted configuration, script / term or style. Next, the correction means corrects the original Japanese sentence by correcting, adding, or deleting the extracted configuration, script, term, or style using data input from the outside. Finally, the output means outputs an easy-to-translate sentence obtained by the correction. With this configuration, points that violate the above rules are extracted and displayed from the original Japanese text, and corrections such as external correction, addition, and deletion are supported.

  Furthermore, another example of a technique related to a document analysis system is disclosed in Non-Patent Document 1 as “prototyping and evaluation of a tool for detecting ambiguity of specifications”. In this ambiguity detection method, the words registered in the dictionary are searched, the ambiguous word candidates are extracted, and the ambiguity level of each ambiguous word candidate is determined according to the usage rules. The step of categorizing into any one makes it possible to selectively detect only words with high ambiguity by excluding unambiguous words and improve the efficiency of correction work.

JP 2007-316834 A

Prototype and evaluation of a tool to detect ambiguity of specifications, Proceedings of IEICE General Conference 2012_Information & Systems, 27, 2012-03-06

  The problem with these disclosed technologies is that even if the analysis method based on the above technique is applied to extract an ambiguous expression from a document written in a natural language, a highly ambiguous part can be detected with high accuracy. It is not. The reason for this is that whether or not an expression that is generally ambiguous is a truly ambiguous expression varies greatly depending on the example in which the expression is used. This is because the method for detecting the presence / absence of an ambiguous word detects an expression that is not ambiguous in each use scene.

  In addition, an expression that is generally ambiguous based on a word and phrase registered in advance in a dictionary, such as that used in the technique of Non-Patent Document 1, is detected, and the example is discriminated by pattern matching with the example. Even with a method that narrows down simple expressions, it is difficult to completely discriminate only those examples that are truly ambiguous with the accuracy of the current natural language processing technology. It is.

  Note that the ambiguity of the document here includes the possibility that at least multiple interpretations will occur, such as whether the writer is difficult to understand between the writer and the reader, and whether the document is easy for the reader to understand. This refers to the characteristics related to the performance of document information transmission.

  That is, the object of the present invention has been described using information as to whether or not the document to be analyzed is written in consideration of ambiguity with respect to the document described in the natural language in view of the above problems. A document analysis system that accurately extracts ambiguous expressions that require preferential correction by changing the conditions for distinguishing only ambiguous examples from ambiguous expressions to reduce false positives, It is to provide a method and program.

A document analysis system according to the present invention is a document analysis system that extracts an ambiguous expression from a document, and a document input unit that receives an input of a target document or document group; and a document that constitutes the document or document group A document analysis unit that extracts word information related to each word used in the document and a portion where the word is used; a document quality is evaluated using a feature of the input document based on a predetermined document quality indexing rule; A document quality evaluation unit that calculates a quality index; character strings of ambiguous words that may have ambiguity, characteristics of examples with different ambiguities for each ambiguous word, and ambiguity that is the degree of ambiguity for each example An ambiguous example database to be stored; based on word information, the ambiguous example database is queried for the presence or absence of ambiguous words in the document. If there are ambiguous words, each word information is An example analysis unit that determines characteristics of an ambiguous word example and queries the ambiguous example database to extract the ambiguous word, its ambiguity, and the position in the document as ambiguous example information about each ambiguous word; A classification accuracy database that collects and accumulates classification accuracy of ambiguous word examples when the ambiguity analysis rule is applied to a document for each ambiguity analysis rule; a document quality index, and an ambiguity analysis rule for a specific word If the classification accuracy obtained from the classification accuracy database for the query is used and the ambiguity analysis rule is a rule that determines whether it is an example with higher ambiguity, the higher the value of the document quality index, the lower the classification accuracy. If the ambiguity analysis rule is a rule that discriminates whether the ambiguity analysis rule is an example with a lower ambiguity, the value of the document quality index is high. Etc., based on a predetermined ambiguity analysis condition change rule to be applied in the classification accuracy poor ambiguity analysis rule, and ambiguity analysis condition changing unit for changing the conditions to analyze the ambiguity for each ambiguous word; changed An ambiguity determination unit that determines the ambiguity of each ambiguous word based on the ambiguity analysis condition; for each ambiguous word that the ambiguity is determined to be equal to or greater than a threshold by the ambiguity determination unit, the corresponding ambiguity and in the document An ambiguous information output unit that outputs an existing position as ambiguous information.

  According to the present invention, it is possible to extract from a document including an ambiguous word written in a natural language to an example having a high ambiguity, thereby limiting to a highly ambiguous portion requiring a preferential correction. It is possible to provide a document analysis system, method, and program capable of pointing out, reducing the load when correcting a document, and improving the efficiency of review.

It is a block diagram which shows the structure of the document analysis system which concerns on embodiment of this invention. It is a sequence diagram which shows the operation example of the document analysis system shown in FIG. It is a block diagram which shows the structure of the document analysis system which concerns on the Example of this invention. It is explanatory drawing which shows the example of a part of ambiguous word Waj, the example Faj, the ambiguity analysis rule Rafj, and the ambiguity Aafj. It is explanatory drawing which shows a part of classification | category precision database regarding the ambiguous word Waj of FIG. 6 is an explanatory diagram illustrating an example of a determination result of an example regarding an ambiguous word Waj having an ambiguity Aafj of 1 or more in FIGS. 6 is an explanatory diagram illustrating an example of a determination result of an example regarding an ambiguous word Waj having an ambiguity Aafj of less than 1 in FIGS.

[Embodiment]
First, an embodiment of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration of a document analysis system 100 according to an embodiment of the present invention.

  Referring to FIG. 1, a document analysis system 100 according to an embodiment of the present invention basically includes at least an electronic device or a system that includes a server and an electronic device and an information communication network such as the Internet that interconnects the server and the electronic device. , Document input unit 10, document analysis unit 20, document quality evaluation unit 30, example analysis unit 40, ambiguity analysis condition change unit 50, ambiguity determination unit 60, ambiguity information output unit 70, and ambiguity An example database 110 and a classification accuracy database 120 are included.

  The illustrated document analysis system 100 uses the degree to which the analysis target document is written in consideration of the ambiguity with respect to the document described in the natural language, and the expression having the described ambiguity. Therefore, it is a document analysis system that accurately extracts ambiguous expressions that need to be preferentially modified by changing the conditions for discriminating only examples that are truly ambiguous to reduce false detections.

  When a document analysis system is configured with electronic devices, the document analysis system 100 can be realized by a computer that operates under program control. Although not shown, this type of computer, as is well known, includes an input device for inputting data, a data processing device, an output device for outputting processing results in the data processing device, and an auxiliary memory serving as various databases. Device. Then, the data processing device stores data in a read-only memory (ROM) that stores a program, a random access memory (RAM) that is used as a work area that temporarily stores data, and a program stored in the ROM. It consists of a central processing unit (CPU) that processes stored data.

  In this case, the input device operates as the document input unit 10, and the data processing device operates as the document analysis unit 20, the document quality evaluation unit 30, the example analysis unit 40, the ambiguity analysis condition change unit 50, and the ambiguity determination unit 60. The auxiliary storage device operates as the ambiguous example database 110 and the classification accuracy database 120, and the output device operates as the ambiguous information output unit 70.

  Next, the operation of each component constituting the document analysis system 100 will be described.

  The document input unit 10 receives an input of a document or a document group for analyzing a highly ambiguous portion that may include ambiguous words and needs to be preferentially corrected.

  The document analysis unit 20 extracts word information of all words used in each sentence by applying morphological analysis to each sentence constituting the document or the document group. Here, in addition to self-supporting words such as nouns, verbs, and adjectives, words are also regarded as individual words such as particles. Also, even if the words are composed of the same character string, if the appearance location is different, word information is extracted for each. Further, the word information includes at least information such as the character string of the word used and the part of speech for each word, the relative positional relationship between the words in the sentence, and the position of the word in the document. The information on the position of the word in the document may be information that can identify the location of use, and includes the order of appearance of the sentence in which the word exists, the page, chapters, sections, and sections on the table of contents.

  The document quality evaluation unit 30 evaluates the document quality using the characteristics of the input document based on a predetermined document quality indexing rule, and calculates a document quality index. Here, the “document quality indexing rule” may be any method that indexes the effectiveness in transmitting the contents of a document to a reader. Examples of document quality indexing include the ratio of sentences with less than a certain number of characters to the total number of sentences in the document, the ratio of sentences in which the subject and predicate correspond one-to-one, and multiple dependency candidates as a result of dependency analysis. An index indicating that the higher the value is, the better the quality of the document, such as the ratio of sentences with no typographical error, the ratio of sentences without typographical errors, or the value that is in a monotonous decrease relationship with the number of writing fluctuations.

  The ambiguous example database 110 is a database that accumulates character strings of ambiguous words that may have ambiguity, characteristics of examples having different ambiguities for each ambiguous word, and ambiguity that is the degree of ambiguity for each example. . The ambiguous example database 110 searches for an inquiry about an arbitrary word and the characteristics of the example of the word to determine whether the word to be inquired matches the ambiguous word stored as a character string, and further matches the ambiguous word Respond to the ambiguity that matches the characteristics of this ambiguous word example.

  Here, the “ambiguity” is an index indicating the degree of ambiguity with respect to the example, and may be an index having a continuous value, or an unclear value such as 0 or 1 indicating whether or not there is ambiguity. It may be an index consisting of continuous values. The “word example” is information in which semantically different usages are classified for each word. Note that a database on the network may be used as the ambiguous example database 110.

  The example analysis unit 40 queries the ambiguous example database 110 for the presence or absence of ambiguous words in the document based on the word information of all words used in each sentence extracted by the document analysis unit 20, and if there is an ambiguous word Based on a predetermined ambiguity analysis rule, the characteristics of each ambiguous word example are determined from the word information, and the ambiguous example database 110 is queried to determine the ambiguous word, its ambiguity, and the position in the document. Extracted as ambiguous example information about each ambiguous word.

  Here, the “ambiguity analysis rule” is a rule for grasping an example of a word from the character string of the word, the part of speech for each word, the relative positional relationship between the words in the sentence, and the like. Use of semantic analysis technology and combination patterns of words and surrounding words is suitable. “Combination pattern of words and surrounding words” refers to a specific position in a sentence in which the word is used, for example, immediately before, immediately after, forward, backward, beginning of a sentence, end of sentence, immediately preceding sentence, immediately following sentence, etc. , Information that classifies whether or not there is predetermined information such as a specific word, part of speech, symbol, numerical expression, or the like.

  The classification accuracy database 120 is a database that collects and accumulates classification accuracy of ambiguous word examples when the ambiguity analysis rule is applied to a document for each ambiguity analysis rule. Then, the classification accuracy database 120 searches and responds to the inquiry of the ambiguity analysis rule regarding the specific word for the classification accuracy.

  Here, “classification accuracy” is an index that indicates the accuracy of separating examples that have different ambiguities while being ambiguous words that consist of the same character string when each ambiguity analysis rule is applied to a document. is there. For example, the classification accuracy may be an index with a continuous value that is applied to an actual document and statistically calculated the accuracy rate of whether the example was correctly separated, or determined based on the experience of the analyst, etc. It may be an index composed of discontinuous values such as high, medium and low. Note that a database on a network may be used as the classification accuracy database 120.

  The ambiguity analysis condition changing unit 50 uses the document quality index calculated by the document quality evaluation unit 30 and the classification accuracy obtained from the classification accuracy database 120 for the query of the ambiguity analysis rule regarding a specific word, Based on the ambiguity analysis condition change rule, the condition for analyzing ambiguity is changed for each ambiguous word.

  Here, the “ambiguity analysis condition change rule” refers to the determination of an example of an ambiguous word so that the higher the value of the document quality index, the more difficult it is to classify it as an example of higher ambiguity. Rules that change conditions are appropriate. In other words, if the ambiguity analysis rule is a rule that determines whether it is an example with a higher ambiguity, the higher the value of the document quality index, the more the ambiguity analysis rule with poor classification accuracy is not applied. In the case of a rule for determining whether or not the example has a lower ambiguity, the higher the value of the document quality index, the more the ambiguity analysis rule with the lower classification accuracy may be applied. For example, ambiguity analysis rules that determine whether or not an example has a high degree of ambiguity are limited to ambiguity analysis rules with a classification accuracy equal to or higher than the classification accuracy threshold set to have a monotonically increasing relationship with the document quality index. For ambiguity analysis rules that discriminate whether or not the usage is low, the ambiguity analysis should be limited to classification accuracy ambiguity analysis rules with a classification accuracy threshold equal to or greater than the document quality index and a monotonic decrease. The method of changing to is suitable.

  In addition, as another example of the ambiguous analysis condition change rule, the classification accuracy of the classification accuracy database 120 is not used, and the determination of whether each example of the ambiguous word is ambiguous is higher as the value of the document quality index is higher. It may be a rule that changes the ambiguity determination condition of an ambiguous word so as to increase the ambiguity determination threshold value that becomes a boundary of the ambiguity that is regarded as ambiguous so that it is difficult to be regarded as ambiguous.

  The ambiguity determination unit 60 determines the ambiguity of each ambiguous word based on the ambiguity analysis condition changed by the ambiguity analysis condition change unit 50 by the ambiguity analysis condition change rule.

  The ambiguity analysis condition change rule is a rule that changes the determination condition of an ambiguous word example so that the higher the document quality index value is, the more difficult it is to classify an example with a higher ambiguity level. Suppose there is. In this case, the ambiguity determination unit 60 performs the example analysis of the ambiguous word again according to the changed analysis condition, and after changing the ambiguity for each example, the ambiguous word of the example having a predetermined value or more is truly ambiguous. Judge as an ambiguous word.

  In addition, regarding the determination of whether the ambiguous analysis condition change rule is ambiguous for each example of an ambiguous word, the higher the value of the document quality index, the more ambiguous the boundary of the ambiguity considered as ambiguous so that it is less likely to be regarded as ambiguous It is assumed that the rule changes the ambiguity determination condition of an ambiguous word so as to increase the determination threshold. In this case, the ambiguity determination unit 60 determines an ambiguous word of an example whose ambiguity is equal to or greater than the ambiguity determination threshold for each example according to the changed analysis condition as a truly ambiguous ambiguous word.

  The ambiguity information output unit 70 outputs the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word that the ambiguity determination unit 60 has determined that the ambiguity is equal to or greater than a threshold value. The output form may be output in a required form.

  For example, the ambiguous information output unit 70 is appropriate to output the entire document by clearly indicating each ambiguous word in the document by color coding, bold emphasis, character enlargement, or the like. In addition, the output form may be a form such as a table in which each ambiguous word is extracted. In addition, as an output form, depending on the degree of ambiguity, color coding, bold emphasis, or the size of a word character may be given.

  Further, the ambiguous information output unit 70 may aggregate the ambiguity for each ambiguous word in the entire document or in an arbitrary range, and output it in a table format as an index representing the quality of the document. Further, the ambiguous information output unit 70 may be able to select each output form so that the display form as a base can be shifted to a table as necessary.

  Next, the overall operation of the document analysis system 100 according to the embodiment of the present invention will be described in detail with reference to the sequence diagrams of FIGS. 1 and 2. Note that the sequence diagram shown in FIG. 2 and the following description are examples of processing, and the processing order and the like may be changed or the processing may be returned or repeated depending on the processing that is appropriately obtained.

  The document input unit 10 receives an input of a document or a document group that analyzes a highly ambiguous part that may contain ambiguous words and that requires preferential correction (step A1 in FIG. 2).

  The document analysis unit 20 extracts word information of all words used in each sentence by applying morphological analysis to each sentence constituting the document or document group (step A2).

  The document quality evaluation unit 30 evaluates the document quality using the characteristics of the input document based on a predetermined document quality indexing rule, and calculates a document quality index (step A3).

  The ambiguous example database 110 accumulates character strings of ambiguous words that may have ambiguity, characteristics of examples having different ambiguities for each ambiguous word, and ambiguity that is the degree of ambiguity for each example. In response to an inquiry about a word and an example feature of the word, a search is made as to whether the query target word matches the ambiguous word stored as a character string, and if the word matches the ambiguous word, the ambiguous word matches the ambiguous word example. The degree is returned (step A4).

  The example analysis unit 40 queries the ambiguous example database 110 for the presence or absence of ambiguous words in the document based on the word information of all words used in each sentence extracted by the document analysis unit 20, and if there is an ambiguous word Based on a predetermined ambiguity analysis rule, the characteristics of each ambiguous word example are determined from the word information, and the ambiguous example database 110 is queried to determine the ambiguous word, its ambiguity, and the position in the document. Extracted as ambiguous example information regarding each ambiguous word (step A5).

  For each ambiguity analysis rule, the classification accuracy database 120 collects and accumulates classification accuracy of ambiguity word examples when the ambiguity analysis rule is applied to a document, and responds to an ambiguity analysis rule query for a specific word. The classification accuracy is searched and responded (step A6).

  The ambiguity analysis condition changing unit 50 uses the document quality index calculated by the document quality evaluation unit 30 and the classification accuracy obtained from the classification accuracy database 120 for the query of the ambiguity analysis rule regarding a specific word, Based on the ambiguity analysis condition change rule, the condition for analyzing ambiguity is changed for each ambiguous word (step A7).

  The ambiguity determination unit 60 determines the ambiguity of each ambiguous word based on the ambiguity analysis condition changed by the ambiguity analysis condition change unit 50 by the ambiguity analysis condition change unit 50 (step A8).

  The ambiguity information output unit 70 outputs the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word determined by the ambiguity determination unit 60 as being ambiguity or more (step A9).

  Next, effects of the document analysis system 100 according to the embodiment of the present invention will be described.

  In the present embodiment, the document features are used, and a document with better document quality is configured to narrow down the ambiguity analysis rule with higher classification accuracy of the example, analyze the example, and calculate the ambiguity. For this reason, when applying uncertain ambiguity analysis rules to documents containing ambiguous words, the analysis conditions can be changed so that the less ambiguous documents are harder to be classified as examples with higher ambiguity. It is possible to reduce false detections when extracting highly ambiguous parts that need correction.

  The document analysis system 100 according to the embodiment of the present invention can be realized as a document analysis method. The document analysis system 100 according to the embodiment of the present invention may be executed by a computer using a document analysis program.

  Next, the operation of the document analysis system 100 according to the embodiment of the present invention will be described using a specific example with reference to FIG.

  The purpose of this embodiment is as follows.

  First, the document analysis system 100 sets an ambiguity IA that is an ambiguity given to the quality of the document D with respect to the ambiguous word Wa in the document D that should exclude an ambiguous part, such as a proposal or a specification regarding information system construction. It is calculated for each usage scene of each ambiguous word Wa, and ambiguity information A regarding the position in the document and the degree of ambiguity of each ambiguous word Wa is estimated. Then, the document analysis system 100 outputs the estimated ambiguity information A, thereby grasping the ambiguity portion to be corrected with priority in the document D and clarifying the low-quality document as a comparison between a plurality of documents. To improve the efficiency of document improvement.

  In the present embodiment, the document analysis system 100 is configured by a document analysis system Y and an intranet server Z as shown in FIG.

  The document analysis system Y operates on the PC terminal possessed by the analyst B, and through the input unit and the output unit, the input of the text constituting the group of documents that the analyst B wants to estimate the ambiguity A is ambiguous. The presentation of degree A is realized.

  The intranet server Z is connected to a PC terminal of an analysis person B who has implemented the document analysis system Y via a communication network. The intranet server Z can search for the presence / absence of registration as an ambiguous word Wa in response to a word query from the document analysis system Y, and can also search for a word example query such as the contents of the example and the number of examples. This is a device that enables retrieval of example information C related to the example.

  Next, the correspondence between FIG. 3 and FIG. 1 will be described.

  The document input unit 10 operates as an input unit of a PC terminal. The document analysis unit 20, the document quality evaluation unit 30, the example analysis unit 40, the ambiguity analysis condition change unit 50, and the ambiguity determination unit 60 are included in the document analysis system Y. The ambiguous information output unit 70 operates as an output unit of the PC terminal. The ambiguous example database 110 and the classification accuracy database 120 are included in the intranet server Z.

  The document analysis system Y and the intranet server Z provided with such means operate as follows.

  The document analysis system Y receives from the input unit an input of a document D that the analysis operator B wants to obtain an ambiguity A such as a proposal or specification regarding information system construction to eliminate an ambiguous part. Then, the document analysis system Y applies morphological analysis to each sentence constituting the document D, decomposes it into words W constituting the document, and all the words Wi (i = 1, 2,... Included in the document D).・ For n), the type of word and the connection relationship between words are extracted as word information. Further, the document analysis system Y assigns sentence numbers as serial numbers in the order of appearance of the sentences in the document, and adds the sentence numbers of sentences including each word to the word information.

  Further, the document analysis system Y counts the number of characters for all the sentences L included in the document D, totals the ratio of sentences having a number of characters less than a certain value as a document quality feature, and calculates the document quality index Nd of the document D. .

  The intranet server Z collects a character string of an ambiguous word Wa that may have ambiguity, an example Fa that changes ambiguity for each ambiguous word Wa, and an ambiguity Aaf that indicates the degree of ambiguity. Accumulated as information Ca. In addition, the intranet server Z also provides a function such as a search engine for extracting information on arbitrary words and expressions, so that the query target word is included in the ambiguous example information Ca in response to a query from the document analysis system Y. It is determined whether it exists, and if it exists, the ambiguity Aaf for each example of the ambiguous word is presented. Note that the ambiguous word to be extracted may be any expression as long as it allows the reader to perform multiple interpretations, such as an abbreviated expression such as “etc.”, “etc.”, an instruction expression such as “that”, “this”, This applies to words such as qualitative expressions such as “large” and “fast”.

  Further, the document analysis system Y inquires the intranet server Z about all the words Wi included in the document D as to whether or not the word corresponds to the ambiguous word in the ambiguous example information Ca, and the determination result indicates that it corresponds to the ambiguous word. All the examples in which the word Wj (j = 1, 2,..., M) is an ambiguous word Waj (j = 1, 2,..., M) and the ambiguity of the ambiguous word Waj changes with respect to the ambiguous word Waj. Extract Fa. Further, the document analysis system Y determines the corresponding example Fa and ambiguity Aaf according to the ambiguity analysis rule Raf prepared for each type of ambiguous word, and determines the ambiguous word Waj and its example Faj, ambiguous. The degree Aafj and the sentence number of an ambiguous word are extracted as ambiguous example information Caj. In addition, when the same ambiguous word is used in the document D several times, it is extracted separately.

  For example, assuming “Principle”, “etc.”, “that”, “rank”, “below”, “other than”, “large” as the ambiguous word Waj, "Make a copy of the file.", "Equal distribution", "If there is a value ...", "5th place", "Following processing", "If less than 10% ...", "A and It is assumed that there are sentences “other than B” and “if there is a value greater than a predetermined value,... In this case, the ambiguity word Waj, the example Faj, the corresponding ambiguity analysis rule Rafj, and the ambiguity Aafj are exemplified as shown in FIG. Here, the ambiguity Aafj is a value determined by an arbitrary method such as a questionnaire. In the case of FIG. 4, the ambiguity is higher as it is larger between 0 and 2, and in the case of 1, the ambiguity of each ambiguous word is standard. If it is smaller than 1, it means that the ambiguity is weaker than the standard. If it is larger than 1, it means that the ambiguity is stronger than the standard.

  Next, for each ambiguity analysis rule Raf, the intranet server Z collects and accumulates the classification accuracy Paf of the example Fa of the ambiguous word Wa when the ambiguity analysis rule Raf is applied to a document as a classification accuracy database. In addition, the intranet server Z also provides a function such as a search engine that extracts the classification accuracy Paf for an arbitrary example Fa, so that an example Fa of an ambiguous word Wa to be queried in response to a query from the document analysis system Y. The categorical accuracy Paf of the ambiguity analysis rule Raf relating to is retrieved and presented.

  For example, the ambiguity analysis rule Raf and the classification accuracy for a part of the example Fa corresponding to “principle”, “etc.”, “that”, “rank”, “below”, “other”, “large” as the ambiguous word Wa Paf is a table as shown in FIG.

Further, the document analysis system Y uses the value of the ambiguity Aafj of each ambiguous word Waj and the document quality index Nd, and classifies the ambiguity analysis rule to be applied as an example discrimination condition according to the following equation (1). A classification accuracy threshold value Ps that is a lower limit of the accuracy Paf is calculated.
Aafj ≧ 1: Ps = Nd
Aafj <1: Ps = 1−Nd (1)

  Further, the document analysis system Y compares the classification accuracy Paf corresponding to the example Faj of each ambiguous word Waj and the classification accuracy threshold value Ps, and applies only the ambiguity analysis rule Raf having the classification accuracy Paf equal to or higher than the classification accuracy threshold value Ps. The determination conditions of the example are changed so that the ambiguity Aafj for each example Faj of each ambiguous word Waj is recalculated and changed.

  This is limited to only an ambiguity analysis rule with a classification accuracy Paf equal to or greater than a classification accuracy threshold proportional to the document quality index Nd when determining whether the ambiguity Aafj is higher than the standard of 1 or more. When determining whether the ambiguity Aafj is lower than the standard of less than 1, the ambiguity analysis rule is limited to the ambiguity analysis rule having the classification accuracy Paf that is smaller than the classification accuracy threshold as the document quality index Nd increases. It means changing the judgment condition.

  For example, when the document quality index Nd is 60% in the examples of FIGS. 4 and 5, the ambiguity analysis rule Rafj of the ambiguous word Waj having the ambiguity Aafj of 1 or more as shown in FIG. Therefore, when the classification accuracy Pafj is 50% [in the same sentence other than “with“ and ”” in front, the ambiguity analysis rule Rafj having an ambiguity Aafj of 1.5 or ““ large ” In the same sentence, the ambiguity analysis rule Rafj having an ambiguity Aafj of 1.2 is changed to inapplicable when determining an example.

  Similarly, in the example of FIGS. 4 and 5, when the document quality index Nd is 60%, the classification accuracy threshold Ps is set in the ambiguity analysis rule Rafj of the ambiguity word Waj having the ambiguity Aafj less than 1 as shown in FIG. Therefore, when the classification accuracy Pafj is 10% [a value immediately before “rank” is a numerical value], the ambiguity analysis rule Rafj having an ambiguity Aafj of 0.5 is not useful in determining an example. Changed to apply.

  Further, the document analysis system Y colors each ambiguous word Wa in the document D based on the ambiguous word Wa, its ambiguity Aafj, and the sentence number of the ambiguous word Waj, thereby clearly identifying the position of the ambiguous sentence to be corrected. Is displayed in an easy-to-understand manner. Further, the document analysis system Y collects the ambiguity Aafj of the ambiguous word Wa based on the ambiguity Aafj in the form of a table or a graph by summing up the entire document D and the chapter unit of the table of contents. It provides metrics that describe the quality of the and the information that determines the ambiguous chapters that need to be corrected.

  The present invention has been described above with reference to the embodiments (and examples), but the present invention is not limited to the above embodiments (and examples). Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

  A part or all of the above-described embodiment can be described as in the following supplementary notes, but is not limited thereto.

(Supplementary note 1) A document analysis system for extracting an ambiguous expression from a document,
A document input unit for receiving input of a target document or document group, and
A document analysis unit for extracting word information about each word used in a document or a sentence constituting a document group and a used part thereof;
A document quality evaluation unit that evaluates document quality using the characteristics of the input document based on a predetermined document quality indexing rule and calculates a document quality index;
An ambiguous word database that accumulates character strings of ambiguous words that may have ambiguity, characteristics of examples with different ambiguities for each ambiguous word, and ambiguity that is the degree of ambiguity for each example;
Based on the word information, the ambiguous word database is queried for the presence or absence of an ambiguous word. If there is an ambiguous word, the characteristic of each ambiguous word example is determined from the word information based on a predetermined ambiguity analysis rule. And, by querying the ambiguous example database, an ambiguous word, its degree of ambiguity, and an existing position in a document are extracted as ambiguous example information about each ambiguous word, respectively,
For each ambiguity analysis rule, a classification accuracy database that collects and accumulates classification accuracy of examples of ambiguous words when the ambiguity analysis rule is applied to a document;
A rule for determining whether the ambiguity analysis rule is an example with higher ambiguity by using the classification accuracy obtained from the classification accuracy database for the document quality index and the query of the ambiguity analysis rule regarding a specific word If so, the higher the value of the document quality index, the more the ambiguity analysis rule with poor classification accuracy is not applied, and the ambiguity analysis rule is a rule for determining whether the ambiguity analysis rule has a lower ambiguity. The higher the value of the document quality index, the more ambiguous the condition for analyzing the ambiguity for each ambiguous word based on the predetermined ambiguity analysis condition change rule that is applied even to the ambiguity analysis rule with poor classification accuracy. Sex analysis condition change part,
An ambiguity determination unit that determines the ambiguity of each ambiguous word based on the changed ambiguity analysis condition;
An ambiguity information output unit that outputs the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word determined by the ambiguity determination unit to be equal to or greater than a threshold;
A document analysis system characterized by comprising:

(Additional remark 2) The ambiguity analysis rule of the said example analysis part is a rule which grasps | ascertains the example of a word from the character string of a word, the part of speech for every word, and the relative positional relationship between the words in a sentence. The document analysis system according to appendix 1, characterized by:

(Supplementary Note 3) The document quality indexing rule of the document quality evaluation unit is a method for indexing effectiveness in transmitting the contents of a document to a reader. Document analysis system.

(Supplementary Note 4) The classification accuracy database collects and accumulates the classification accuracy of a document group on the intranet as an example of the ambiguous word or a document group of the same domain as the document to be analyzed, The document analysis system according to any one of supplementary notes 1 to 3.

(Additional remark 5) Regarding the ambiguity analysis condition change rule of the ambiguity analysis condition change unit, regarding the discrimination of the examples having different ambiguities, the higher the value of the document quality index, the more difficult it is to classify the ambiguity analysis condition change rule. The document analysis system according to any one of appendices 1 to 4, wherein the document analysis system is a rule for changing a determination condition of an ambiguous word example.

(Additional remark 6) The said ambiguity determination part implements the example analysis of an ambiguous word again according to the changed analysis conditions, and after changing the ambiguity for every example, an ambiguous word of an example more than a predetermined value is truly ambiguous. The document analysis system according to appendix 5, wherein the document analysis system is determined as an ambiguous word.

(Additional remark 7) Regarding the determination of whether the ambiguity analysis condition change rule of the ambiguity analysis condition change unit is ambiguous for each example of an ambiguous word, the higher the value of the document quality index, the more ambiguous it is The document analysis according to any one of appendices 1 to 4, which is a rule for changing an ambiguity determination condition of an ambiguous word so as to increase an ambiguity determination threshold that is a boundary of ambiguity to be regarded as system.

(Additional remark 8) The said ambiguousness determination part determines the ambiguous word of the example whose ambiguity is more than the said ambiguity determination threshold value for every example according to the changed analysis conditions as a truly ambiguous ambiguous word. 8. The document analysis system according to 7.

(Supplementary note 9) A document analysis method for extracting an ambiguous expression from a document,
A document input step in which the document input unit accepts input of a target document or document group; and
A document analysis step in which a document analysis unit extracts each word used in a sentence that constitutes a document or a document group and word information related to the use location;
A document quality evaluation unit that evaluates document quality using the characteristics of the input document based on a predetermined document quality indexing rule and calculates a document quality index; and
Based on the word information, the example analysis unit determines whether or not there is an ambiguous word in the document. The character string of the ambiguous word that may have ambiguity and the characteristics of the example with different ambiguity for each ambiguous word and the ambiguity for each example Queries the ambiguity example database that accumulates the degree of ambiguity, and if there is an ambiguity word, based on a predetermined ambiguity analysis rule, the feature of each ambiguity word example is determined from the word information, and the ambiguity example An example analysis step of extracting an ambiguous word, its ambiguity, and an existing position in a document as ambiguous example information for each ambiguous word by querying the database,
An ambiguous word when the ambiguity analysis condition changing unit applies the ambiguity analysis rule to a document for each ambiguity analysis rule in response to an inquiry of the ambiguity analysis rule regarding the document quality index and a specific word If the classification accuracy obtained from the classification accuracy database that collects and accumulates the classification accuracy of the example is used and the ambiguity analysis rule is a rule that determines whether it is an example with a higher ambiguity, the value of the document quality index The higher the value of the document quality index is, the higher the value of the document quality index is, the higher the value of the document quality index is, the higher the value of the document quality index is. based on a predetermined ambiguity analysis condition change rule to be applied in the classification accuracy poor ambiguity analysis rule, varying the conditions to analyze the ambiguity for each ambiguous word And ambiguity analysis conditions changing step of,
An ambiguity determination unit determines an ambiguity of each ambiguous word based on the changed ambiguity analysis condition,
The ambiguous information output unit outputs the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word for which the ambiguity is determined to be greater than or equal to the threshold in the ambiguity determination step,
A document analysis method comprising:

(Additional remark 10) The said ambiguity analysis rule is a rule which grasps | ascertains the example of a word from the character string of a word, the part of speech for every word, and the relative positional relationship between the words in a sentence, The document analysis method according to attachment 9.

(Supplementary note 11) The document analysis method according to supplementary note 9 or 10, wherein the document quality indexing rule is a method for indexing the effectiveness in transmitting the contents of a document to a reader.

(Additional remark 12) Regarding the determination of examples with different ambiguity, the ambiguity analysis condition change rule is such that the higher the value of the document quality index is, the more difficult it is to classify the ambiguity example as a high ambiguity example. 12. The document analysis method according to any one of appendices 9 to 11, wherein the document analysis method is a rule for changing an item.

(Additional remark 13) In the said ambiguity determination step, the said ambiguity determination part re-executes the example analysis of an ambiguous word according to the changed analysis conditions, changes an ambiguity for every example, and is an example more than a predetermined value The document analysis method according to appendix 12, wherein the ambiguous word is determined as a truly ambiguous ambiguous word.

(Supplementary Note 14) Regarding the determination of whether or not the ambiguity analysis condition change rule is ambiguous for each example of an ambiguous word, the higher the value of the document quality index, the less likely it is to be regarded as ambiguous. The document analysis method according to any one of appendices 9 to 11, wherein the ambiguity determination condition is changed so as to increase the ambiguity determination threshold.

(Supplementary Note 15) In the ambiguity determination step, the ambiguity determination unit determines an ambiguous word of an example whose ambiguity is equal to or greater than the ambiguity determination threshold for each example according to the changed analysis condition as a truly ambiguous ambiguous word. The document analysis method according to appendix 14, characterized by:

(Supplementary Note 16) A document analysis program for causing a computer to extract an ambiguous expression from a document, wherein the computer
A document input procedure for receiving input of a target document or document group,
A document analysis procedure for extracting word information related to each word used in a document or a sentence constituting a document group and a used part thereof;
A document quality evaluation procedure for evaluating the document quality using the characteristics of the input document based on a predetermined document quality indexing rule and calculating the document quality index;
Based on the word information, whether or not there is an ambiguous word in the document is a character string of an ambiguous word that may have ambiguity and an example feature with different ambiguity for each ambiguous word and the degree of ambiguity for each example. Queries in the ambiguous example database that accumulates the ambiguity, if there are ambiguous words, based on a predetermined ambiguity analysis rule, determine the characteristics of the example of each ambiguous word from the word information, by querying the ambiguous example database , An example analysis procedure for extracting ambiguous words, their ambiguity, and their position in the document as ambiguous example information for each ambiguous word,
In response to the query of the ambiguity analysis rule regarding the document quality index and the specific word, the classification accuracy of the ambiguity word example when the ambiguity analysis rule is applied to the document for each ambiguity analysis rule is collected. If the classification accuracy obtained from the classification accuracy database stored in the table is used and the ambiguity analysis rule is a rule for determining whether the ambiguity analysis rule is a higher ambiguity example, the higher the value of the document quality index, the higher the classification accuracy. A bad ambiguity analysis rule is not applied, and if the ambiguity analysis rule is a rule for determining whether the ambiguity analysis rule is an example with a lower ambiguity, the higher the value of the document quality index, the worse the ambiguity analysis with the classification accuracy. based on a predetermined ambiguity analysis condition change rule to be applied in the rule, ambiguity analysis condition change for changing the conditions to analyze the ambiguity for each ambiguous word And order,
An ambiguity determination procedure for determining the ambiguity of each ambiguous word based on the changed ambiguity analysis condition;
An ambiguity information output procedure for outputting the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word determined to have an ambiguity equal to or greater than a threshold in the ambiguity determination step;
Document analysis program that executes

(Additional remark 17) The said ambiguity analysis rule is a rule which grasps | ascertains the example of a word from the character string and the part of speech for every word, and the relative positional relationship between the words in a sentence, It is characterized by the above-mentioned. The document analysis program according to attachment 16.

(Supplementary note 18) The document analysis program according to supplementary note 16 or 17, characterized in that the document quality indexing rule is a method for indexing the effectiveness in transmitting the contents of a document to a reader.

(Additional remark 19) Regarding the determination of the ambiguity analysis condition change rule, the determination condition of the ambiguity word example is such that the higher the value of the document quality index, the more difficult it is to classify the ambiguity analysis condition change rule. 19. The document analysis program according to any one of supplementary notes 16 to 18, wherein

(Additional remark 20) The said ambiguity determination procedure performs the example analysis of an ambiguous word again according to the changed analysis conditions, and after changing the ambiguity for every example, an ambiguous word of an example more than a predetermined value is truly ambiguous. The document analysis program according to appendix 19, wherein the document analysis program is determined as an ambiguous word.

(Additional remark 21) Regarding the determination of whether or not the ambiguity analysis condition change rule is ambiguous for each example of an ambiguous word, the higher the value of the document quality index, the more ambiguous the boundary that is regarded as ambiguous. The document analysis program according to any one of appendices 16 to 18, wherein the ambiguity determination condition is changed so as to increase the ambiguity determination threshold.

(Additional remark 22) The said ambiguousness determination procedure determines the ambiguous word of the example whose ambiguity is more than the said ambiguity determination threshold value for every example according to the changed analysis conditions as a truly ambiguous ambiguous word. 21. The document analysis program according to 21.

  According to the present invention, it is possible to preferentially correct the ambiguity of various documents exchanged in work such as requirement definition in software and system development, particularly from a place where a problem is large, This leads to the efficiency of document review, and the situation where different interpretations occur among multiple readers is reduced. Therefore, it can be applied to applications related to the efficiency of system development, such as reducing rework and improving customer satisfaction.

DESCRIPTION OF SYMBOLS 10 Document input part 20 Document analysis part 30 Document quality evaluation part 40 Example analysis part 50 Ambiguity analysis condition change part 60 Ambiguity determination part 70 Ambiguous information output part 110 Ambiguous example database 120 Classification accuracy database D Document A Ambiguous word Y Document analysis System Z intranet server

Claims (10)

A document analysis system that extracts ambiguous expressions from a document,
A document input unit for receiving input of a target document or document group, and
A document analysis unit for extracting word information about each word used in a document or a sentence constituting a document group and a used part thereof;
A document quality evaluation unit that evaluates document quality using the characteristics of the input document based on a predetermined document quality indexing rule and calculates a document quality index;
An ambiguous word database that accumulates character strings of ambiguous words that may have ambiguity, characteristics of examples with different ambiguities for each ambiguous word, and ambiguity that is the degree of ambiguity for each example;
Based on the word information, the ambiguous word database is queried for the presence or absence of an ambiguous word. If there is an ambiguous word, the characteristic of each ambiguous word example is determined from the word information based on a predetermined ambiguity analysis rule. And, by querying the ambiguous example database, an ambiguous word, its degree of ambiguity, and an existing position in a document are extracted as ambiguous example information about each ambiguous word, respectively,
For each ambiguity analysis rule, a classification accuracy database that collects and accumulates classification accuracy of examples of ambiguous words when the ambiguity analysis rule is applied to a document;
A rule for determining whether the ambiguity analysis rule is an example with higher ambiguity by using the classification accuracy obtained from the classification accuracy database for the document quality index and the query of the ambiguity analysis rule regarding a specific word If so, the higher the value of the document quality index, the more the ambiguity analysis rule with poor classification accuracy is not applied, and the ambiguity analysis rule is a rule for determining whether the ambiguity analysis rule has a lower ambiguity. The higher the value of the document quality index, the more ambiguous the condition for analyzing the ambiguity for each ambiguous word based on the predetermined ambiguity analysis condition change rule that is applied even to the ambiguity analysis rule with poor classification accuracy. Sex analysis condition change part,
An ambiguity determination unit that determines the ambiguity of each ambiguous word based on the changed ambiguity analysis condition;
An ambiguity information output unit that outputs the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word determined by the ambiguity determination unit to be equal to or greater than a threshold;
A document analysis system characterized by comprising:   The ambiguity analysis rule of the example analysis unit is a rule for grasping an example of a word from a character string of a word, a part of speech for each word, and a relative positional relationship between words in a sentence. The document analysis system according to claim 1.   3. The document analysis system according to claim 1, wherein the document quality indexing rule of the document quality evaluation unit is a method for indexing effectiveness in transmitting the contents of a document to a reader. . The classification accuracy database collects and accumulates the classification accuracy of a document group on an intranet as an example of the ambiguous word or a document group of the same domain as a document to be analyzed. 4. The document analysis system according to any one of items 1 to 3.   The ambiguity analysis condition change rule of the ambiguity analysis condition change unit is an example of an ambiguous word so that the higher the value of the document quality index, the more difficult it is to classify the ambiguity analysis condition change rule. 5. The document analysis system according to claim 1, wherein the rule is to change the determination condition.   The ambiguity determination unit performs the example analysis of the ambiguous word again according to the changed analysis condition, and after changing the ambiguity for each example, the ambiguous word of the example exceeding the predetermined value is regarded as a truly ambiguous ambiguous word. 6. The document analysis system according to claim 5, wherein the determination is performed.   The degree of ambiguity that the ambiguity analysis condition change rule of the ambiguity analysis condition change unit regards as ambiguous so that the higher the value of the document quality index, the less likely it is to be regarded as ambiguous as to whether or not each example of an ambiguous word is ambiguous 5. The document analysis system according to claim 1, wherein the ambiguity determination condition of an ambiguous word is changed so as to increase an ambiguity determination threshold value that becomes a boundary of the document.   8. The ambiguity determination unit determines an ambiguity word of an example whose ambiguity is equal to or greater than the ambiguity determination threshold for each example according to the changed analysis condition as a truly ambiguous ambiguity word. Document analysis system. A document analysis method for extracting an ambiguous expression from a document,
A document input step in which the document input unit accepts input of a target document or document group; and
A document analysis step in which a document analysis unit extracts each word used in a sentence that constitutes a document or a document group and word information related to the use location;
A document quality evaluation unit that evaluates document quality using the characteristics of the input document based on a predetermined document quality indexing rule and calculates a document quality index; and
Based on the word information, the example analysis unit determines whether or not there is an ambiguous word in the document. Queries the ambiguity example database that accumulates the degree of ambiguity, and if there is an ambiguity word, based on a predetermined ambiguity analysis rule, the feature of each ambiguity word example is determined from the word information, and the ambiguity example An example analysis step of extracting an ambiguous word, its ambiguity, and an existing position in a document as ambiguous example information for each ambiguous word by querying the database,
An ambiguous word when the ambiguity analysis condition changing unit applies the ambiguity analysis rule to a document for each ambiguity analysis rule in response to an inquiry of the ambiguity analysis rule regarding the document quality index and a specific word If the classification accuracy obtained from the classification accuracy database that collects and accumulates the classification accuracy of the example is used and the ambiguity analysis rule is a rule that determines whether it is an example with a higher ambiguity, the value of the document quality index The higher the value of the document quality index is, the higher the value of the document quality index is, the higher the value of the document quality index is, based on a predetermined ambiguity analysis condition change rule to be applied in the classification accuracy poor ambiguity analysis rule, varying the conditions to analyze the ambiguity for each ambiguous word And ambiguity analysis conditions changing step of,
An ambiguity determination unit determines an ambiguity of each ambiguous word based on the changed ambiguity analysis condition,
The ambiguous information output unit outputs the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word for which the ambiguity is determined to be greater than or equal to the threshold in the ambiguity determination step,
A document analysis method comprising: A document analysis program for causing a computer to extract an ambiguous expression from a document, wherein the computer
A document input procedure for receiving input of a target document or document group,
A document analysis procedure for extracting word information related to each word used in a document or a sentence constituting a document group and a used part thereof;
A document quality evaluation procedure for evaluating the document quality using the characteristics of the input document based on a predetermined document quality indexing rule and calculating the document quality index;
Based on the word information, whether or not there is an ambiguous word in the document is a character string of an ambiguous word that may have ambiguity and an example feature with different ambiguity for each ambiguous word and the degree of ambiguity for each example. Queries in the ambiguous example database that accumulates the ambiguity, if there are ambiguous words, based on a predetermined ambiguity analysis rule, determine the characteristics of the example of each ambiguous word from the word information, by querying the ambiguous example database , An example analysis procedure for extracting ambiguous words, their ambiguity, and their position in the document as ambiguous example information for each ambiguous word,
In response to the query of the ambiguity analysis rule regarding the document quality index and the specific word, the classification accuracy of the ambiguity word example when the ambiguity analysis rule is applied to the document for each ambiguity analysis rule is collected. If the classification accuracy obtained from the classification accuracy database stored in the table is used and the ambiguity analysis rule is a rule for determining whether the ambiguity analysis rule is a higher ambiguity example, the higher the value of the document quality index, the higher the classification accuracy. A bad ambiguity analysis rule is not applied, and if the ambiguity analysis rule is a rule for determining whether the ambiguity analysis rule is an example with a lower ambiguity, the higher the value of the document quality index, the worse the ambiguity analysis with the classification accuracy. based on a predetermined ambiguity analysis condition change rule to be applied in the rule, ambiguity analysis condition change for changing the conditions to analyze the ambiguity for each ambiguous word And order,
An ambiguity determination procedure for determining the ambiguity of each ambiguous word based on the changed ambiguity analysis condition;
An ambiguity information output procedure for outputting the corresponding ambiguity and the position in the document as ambiguity information for each ambiguous word determined to have an ambiguity equal to or greater than a threshold in the ambiguity determination step;
Document analysis program that executes

Images