JP2010117832A - Related information extraction device, related information extraction method, program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device and a method for highly precisely extracting related information among a plurality of unique representations corresponding to each case, and to provide a program and a recording medium. <P>SOLUTION: This related information extraction device for extracting information about a plurality of input unique representations includes: an analysis processing part 10 for, when a text including each unique representation is input, performing the morphemic analysis of the input text, and for analyzing the dependency of clauses configuring an input text; and a related information extraction processing part 20 for, when acquiring an analytic result by the analysis processing part 10, extracting at least one independent word included in the input text as a related information candidate, and for acquiring related estimation information showing the degree of estimation that the related information candidate is related information for each extracted related information candidate, and for extracting related information from the related information candidates on the basis of the analytic result and the related estimation information. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a technique for extracting information related to a plurality of specific expressions, which plays an important role in a summarization system or the like that summarizes input text.

  Conventionally, as a technique for extracting information related to a plurality of specific expressions, a large number of sentences including two specific expressions are stored in a predetermined storage device in advance, and words existing between the specific expressions for these sentences or It is known to search for words existing within a predetermined number of characters before and after each unique expression and extract the most frequently searched words as related information (see Non-Patent Document 1, for example).

  As shown in FIG. 1, such a relationship information extraction device includes a storage unit 1, a text acquisition unit 2, a target extraction unit 3, and a relationship information extraction unit 4.

  The storage unit 1 stores in advance a plurality of sentences that have been subjected to a known morphological analysis process. The text acquisition unit 2 acquires a sentence including two specific expressions input using an input unit such as a keyboard from the storage unit. For example, when two unique expressions “Koizumi” and “Japan” are input, the text acquisition unit 2 reads “Koizumi, the prime minister of Japan, will visit Korea next month.” The Prime Minister visited the US last month and met with US President Bush. ” In this case, the sentence acquired from the storage unit is “Japan (noun) / no (case particle) / prime (noun) / de (particle) / ar (verb) / Koizumi (noun) / Mr. (Suffix) / Is (participant) / next month (noun) / Korea (noun) / do (case particle) / visit (verb) / do (suffix) /. (Phrase) ”or“ Japan (noun) / no (case particle) ” / Koizumi (noun) / prime minister (noun) / ha (particle) / last month (noun) / visiting the United States (verb) / shi (suffix) / bush (noun) / rice (noun) / president (noun) / and (particle) ) / Conversation (verb) / done (suffix) /. (Punctuation) ”and so on.

  The target extraction unit 3 uses the two specific expressions to extract words existing between the specific expressions and words existing within 10 characters before and after each specific expression from the sentence acquired by the text acquisition unit 2. . Here, using each extracted sentence as an example, “Prime” and “A” are extracted as words existing between each unique expression, and words existing within 10 characters before and after each specific expression. “Next month”, “Korea”, “Visit”, “Prime Minister”, “Last month”, “Visit America” and “Bush” are extracted.

  Next, the target extraction unit 3 counts the number of extractions of each extracted word and outputs each word and the number of extractions. If it explains using the above-mentioned extraction result, the word "Prime Minister" is extracted twice, and the other words are extracted once each.

  When the relationship information extraction unit 4 acquires each word and the number of extractions from the target extraction unit 3, the relationship information extraction unit 4 outputs the word with the highest number of extractions as the relationship information. In this case, the word “Prime Minister” is output as the relationship information.

  In this way, the word “Prime Minister” is extracted as the relation information of the two unique expressions “Koizumi” and “Japan”.

  In the conventional relationship information extraction process, for example, it is possible to improve the extraction accuracy of relationship information by storing in advance in the storage unit 1 a large number of texts including two specific expressions “Koizumi” and “Japan”, Therefore, although it is possible to extract information representing potential relationships such as “Prime Minister” and “Prime Minister” for the two unique expressions “Koizumi” and “Japan”, for example, “Junichiro Koizumi gave a speech at Tokyo Station It was difficult to extract information representing the temporary relationship between “Joiichiro Koizumi” and “Tokyo Station” in the case of

  In view of this, the present inventor has already proposed a relationship information extraction apparatus and method capable of extracting information representing a temporary relationship with respect to each unique information (see, for example, Patent Document 1).

  When the text including each unique expression is input, the relation information extracting device analyzes the input text by morphological analysis and analyzes the dependency of the clauses constituting the input text, and the analysis result by the analysis processing section. In addition, when the specific expressions are included in the same sentence, the independent words included in the unrelated clauses are extracted as relation information in the dependency analysis result between the clauses including the specific expressions. If each specific expression is contained in a different sentence, the relation information that extracts independent words different from the one specific expression from the clause containing one specific expression as the related information. And an extraction processing unit.

As a result, it is an independent word included in the clause constituting the input text, and one of the independent word or each unique expression included in the unrelated clause in the dependency analysis result between the phrases including each specific expression. Since the independent words in the phrase including the specific expressions are extracted as the relationship information, the relationship information between the specific expressions can be extracted from the input text.
Junichiro Mori and three others, “Extraction of relationship information between entities from the Web”, [online], December 2006, Japan Society for Artificial Intelligence, [searched on January 22, 2007], Internet <URL: http : //www.jstage.jst.go.jp/article/pjsai/JSAI06/0/12/#pdf/-char/en/> JP 2008-225566 A

  However, in the technique described in Patent Literature 1, for example, in the case of “Junichiro Koizumi made a speech at Tokyo Station”, information representing a temporary relationship “Speech” with respect to “Junichiro Koizumi” and “Tokyo Station” was extracted. However, it is difficult to extract the relationship information with high accuracy because the evaluation result that the extraction accuracy of the optimum relationship information is 19% and the recall rate of the relationship information in the same case is 33% is obtained. Met.

  The present invention has been made in view of the above-mentioned problems, and an object of the present invention is an apparatus, a method, a program, and an apparatus capable of extracting relation information between a plurality of specific expressions according to individual cases with high accuracy. It is to provide a recording medium.

  In order to achieve the above object, the related information extracting apparatus of the present invention is an apparatus for extracting information related to a plurality of input specific expressions, and when a text including each of the specific expressions is input, When the analysis processing unit that analyzes the morphological analysis of the text and analyzes the dependency of the clauses constituting the input text, and when the analysis result by the analysis processing unit is acquired, at least one independent word included in the input text is extracted as a related information candidate In addition, relation estimation information indicating the degree to which the relation information candidate is estimated to be relation information is acquired for each extracted relation information candidate, and relation information is extracted from the relation information candidate based on the analysis result and the relation estimation information And a related information extraction processing unit.

  Further, the related information extraction method of the present invention is a method for extracting information related to a plurality of input specific expressions using a computer in order to achieve the object, wherein the computer Input text, morphological analysis of the input text, analysis of the dependency of clauses constituting the input text, and extraction of at least one independent word included in the input text as a candidate for related information The relationship estimation information indicating the degree to which the information candidate is estimated to be the relationship information is acquired for each extracted relationship information candidate, and the relationship information is extracted from the relationship information candidate based on the analysis result and the relationship estimation information.

  Furthermore, the program of the present invention is for causing a computer to function as each means of the related information extracting apparatus.

  Furthermore, the program of the present invention is for causing a computer to execute each process of the related information extraction method.

  Furthermore, the recording medium of the present invention records the above program.

  As a result, at least one independent word included in the input text is extracted as a related information candidate, and the related information from the related information candidate based on the relationship estimation information and the analysis result indicating the degree to which the related information candidate is estimated to be related information. Since the information is extracted, it is possible to extract the relationship information between each unique expression from the input text.

  According to the related information extracting apparatus, method, program and recording medium of the present invention, the related information of each unique expression can be extracted from the input text. For example, “Junichiro Koizumi is Tokyo Station and Ichiro Ozawa is Osaka Station In the case of "Speech", the information representing the temporary relationship "Speech" can be extracted for the specific expressions "Joiichiro Koizumi" and "Tokyo Station". Related information can be extracted.

  In addition, according to the present invention, the extraction accuracy of the optimum relationship information is greatly improved from 19% to 68% as compared with the technique described in Japanese Patent Application Laid-Open No. 2008-225565 described above, and the relationship in the same case The special effect that the information reproduction rate was improved from 33% to 41% was obtained.

  2 to 9 show an embodiment of the present invention. FIG. 2 is a configuration diagram of a related information extracting apparatus according to an embodiment of the present invention. FIG. 3 is a flowchart of related information extraction processing. FIG. 5 is a diagram showing an overview of the analysis result by the receiving analysis unit, FIG. 5 is a diagram showing an overview of the processing result by the specific expression matching unit, and FIGS. 6 to 8 are diagrams showing an overview of the processing result by the dependency structure information acquiring unit, FIG. 9 is a diagram illustrating an outline of a processing result by the relationship estimation information acquisition unit.

  The outline of the related information extracting apparatus and method according to the present invention will be described below with reference to the drawings.

  The related information extracting apparatus of the present invention comprises a computer device mainly composed of a well-known CPU, and can be connected to display means such as a monitor, input means such as a keyboard, storage means such as a hard disk and memory, and an external network. A device (not shown) is provided. Further, the relationship information extraction apparatus of the present invention is provided with an analysis processing unit 10, a relationship information extraction processing unit 20, a relationship estimation information storage unit 30, and a model storage unit 40.

  As shown in FIG. 2, the analysis processing unit 10 includes a morphological analysis unit 11, a dependency analysis unit 12, and a specific expression association unit 13. The analysis processing unit 10 performs morphological analysis on the text input using the input unit and configures the input text. It is designed to analyze the dependency relations of phrases.

  When the morphological analysis unit 11 acquires the input text (step S1 in FIG. 3), the input text is subjected to a known morphological analysis process to divide the input text into words, and parts of speech are assigned to the divided words. Output (step S2 in FIG. 3). For example, if a sentence “Junichiro Koizumi gave a speech at Tokyo Station and Ichiro Ozawa at Osaka Station” is input, the processing result by the morphological analysis unit 11 is “Junichiro Koizumi (noun) / ha (case particle). ) / Tokyo Station (noun) / de (particle) /, (reading) / Ichiro Ozawa (noun) / ha (particle) / Osaka station (noun) / de (particle) / speech (verb) / suffix (suffix) /. (Punctuation) ".

  The dependency analysis unit 12 divides the text into phrases by performing a well-known dependency analysis process on the input text that has been obtained from the morpheme analysis unit 11 and has been subjected to morpheme analysis. The dependency relationship is analyzed and output (step S3 in FIG. 3). In this case, when the illustrated input text is analyzed by the dependency analysis unit 12, information indicating a dependency structure as shown in FIG. 4 (dependency tree) is output as an analysis result. Here, the phrase “Junichiro Koizumi / Ha”, the phrase “Tokyo Station / De /,”, the phrase “Ichiro Ozawa / Ha”, and the phrase “Osaka Station / De” are respectively “Speech / In the case of implementing these dependency relationships as data, for example, “(Speech. (Junichiro Koizumi) (At Tokyo Station)” (Ichiro Ozawa) (Osaka) In addition, these dependency relationships include dependency type information defined in a well-known dependency analysis technique (“D” representing a normal dependency relationship, parallel dependency). "P" representing a receiving relationship or "A" representing a similar dependency relationship) is assigned.

  The specific expression association unit 13 is for acquiring a set of specific expressions composed of two specific expressions and performing a specific expression association process using the analysis result of the dependency analysis unit 12. More specifically, when the specific expression association unit 13 obtains a set of specific expressions input using the input unit and the input text analyzed by the dependency analysis unit 12 (step S4 in FIG. 3), the input is performed. A unique expression corresponding to each unique expression is extracted from the input text, and a unique expression identifier representing the specific expression is assigned to the extracted specific expression (step S5 in FIG. 3). For example, when a specific expression “Junichiro Koizumi” is input, “Junichiro Koizumi” in the input text is expressed as “<PSN> Junichiro Koizumi </ PSN>”. Here, “PSN” is a unique expression identifier representing a person name. In this embodiment, eight types of unique expression identifiers (“PSN” representing a person name and “PSN” representing an organization name defined in a well-known unique expression extraction technique). ORG ”,“ LOC ”representing place name,“ ART ”representing unique name,“ MNY ”representing amount,“ PNT ”representing percentage,“ TIM ”representing time, and“ DAT ”representing date) ing.

  In the present embodiment, a set of unique expressions is expressed as “Junichiro Koizumi: Tokyo Station”. In this case, the specific expression that appears first in the input text in the set of specific expressions is represented as the front specific expression on the left side of “:”, and the specific expression that appears later is displayed on the right side of “:” as the backward specific expression. Further, when a set of specific expressions “Junichiro Koizumi: Tokyo Station” is input, the processing result of the specific expression association unit 13 is shown in FIG.

  Next, an outline of the relationship information extraction processing unit 20 will be described. The relationship information extraction processing unit 20 includes a relationship information candidate extraction unit 21, a dependency structure information acquisition unit 22, a relationship estimation information acquisition unit 23, a model selection unit 24, a classifier 25, and a relationship information extraction unit 26. The relation information of the set of specific expressions is extracted based on the analysis result from the analysis processing unit 10.

  When the analysis result obtained by the analysis processing unit 10 is acquired from the specific expression association unit 13, the related information candidate extraction unit 21 extracts at least one independent word included in the input text as a related information candidate (step S6 in FIG. 3). . Here, the relationship information candidate extraction unit 21, when a set of specific expressions is included in the same phrase in the input text, between the independent words adjacent to the front of the front specific expression in the specific expressions and each specific expression The existing independent words and the independent words that are adjacent to the rear of the backward unique expression are extracted as related information candidates. For example, when a set of specific expressions “Ishihara: Tokyo” is included in the same phrase, such as “<PSN> Ishihara </ PSN> <LOC> Tokyo </ LOC> governor”, An independent word “governor” adjacent to the back of the backward proper expression “Tokyo” is extracted as a candidate for related information.

  In addition, the related information candidate extraction unit 21, when the set of specific expressions in the input text is included in different clauses, the independent word adjacent to the front or rear of the front specific expression in the set of specific expressions, and the rear specific expression Independent words adjacent to the front or rear of the are extracted as related information candidates. For example, if the set of specific expressions “Joiichiro Koizumi: Liberal Democratic Party” is included in different clauses, such as “<ORG> Liberal Democratic Party </ ORG> Governors”, the backward proper expression “Liberal Democratic Party” An independent word “Governor” adjacent to the rear is extracted as a related information candidate.

  Furthermore, when the main word of the phrase that does not include the specific expression pair is an independent word, the related information candidate extraction unit 21 extracts the morphemes from the head of the phrase to the main word as related information candidates. For example, when the main sentence “Speech” is an independent word as in the phrase “Speech”, the morpheme “Speech” is extracted as a related information candidate.

  In the present embodiment, when the analysis result shown in FIG. 5 is input, “Ichiro Ozawa”, “Osaka Station”, and “Speech” are extracted as related information candidates. Further, in the present embodiment, the above three relational information candidate extraction methods are used. However, any one or two of the relational information candidate extraction methods may be used to perform the relational information candidate extraction process. .

  When the dependency structure information acquisition unit 22 acquires the analysis result by the analysis processing unit 10 from the specific expression association unit 13 and acquires the related information candidate from the related information candidate extraction unit 21, the set of specific expressions and the related information candidate Extract the smallest dependency tree that contains. In addition, the dependency structure information acquisition unit 22 includes the extracted dependency tree, the part of speech / dependency type / participant of the main part of each clause, the specific expression identifier of the set of specific expressions, the notation of the morphemes of the related information candidates, The dependency structure information is acquired by generating the tree structure information using the part of speech (step S7 in FIG. 3). For example, the tree structure information generated from the unique expression set “Junichiro Koizumi: Tokyo Station” and the related information candidate “Ichiro Ozawa” is shown in FIG. Further, the tree structure information generated from the unique expression set “Junichiro Koizumi: Tokyo Station” and the related information candidate “Osaka Station” is shown in FIG. 7 and is called “Junichiro Koizumi: Tokyo Station”. The tree structure information generated from the set of proper expressions and the related information candidate “speech” is shown in FIG. In the present embodiment, “φ” is used to express that the particle is not included in the phrase. For example, the tree structure information in FIG. 6 to FIG. 8 indicates that the particle is not included in the phrase “spoken”.

  When the relationship estimation information acquisition unit 23 acquires the dependency structure information of each relationship information candidate generated by the dependency structure information acquisition unit 22, the relationship estimation information indicating the degree to which the relationship information candidate is estimated to be the relationship information. Is acquired from the relationship estimation information storage unit 30 (step S8 in FIG. 3). The relationship estimation information storage unit 30 has a plurality of relationship estimation information corresponding to each of a plurality of morphemes, for example, “Ichiro Ozawa = 0.1”, “Osaka Station = 0.2”, “Speech = 1”. Each relationship estimation information is stored in advance, and is created by a human or a calculation method described later. When the relationship estimation information acquisition unit 23 acquires the dependency structure information illustrated in FIG. 8, among the relationship estimation information stored in the relationship estimation information storage unit 30, the relationship estimation corresponding to the relationship information candidate “speech”. Information is searched, and relevant relationship estimation information “1” is extracted. Then, the relationship estimation information acquisition unit 23 adds the extracted relationship estimation information as a child node of the candidate node in the dependency structure information as illustrated in FIG. The relationship estimation information acquisition unit 23 performs the same process on the other relationship information candidates (“Ichiro Ozawa” and “Osaka Station”) extracted by the relationship information candidate extraction unit 21.

  As for the technique for calculating the estimated degree of related information from a large corpus, Tanaka et al., “Classification of syntactic categories of nouns based on the dispersion of semantic categories”, Transactions of Information Processing Society of Japan, vol. 40, no. 9, pp. 3387-3396, September 1999.

  When the model selection unit 24 acquires the processing result of the relationship estimation information acquisition unit 23, the model selection unit 24 classifies the set of proper expressions based on the specific expression identifier assigned by the specific expression association unit 13, and uses a classifier 25 described later. The type of model to be extracted is selected (step S9 in FIG. 3). For example, when “Joiichiro Koizumi: Tokyo Station” is input as a set of specific expressions, the model selection unit 24 classifies the set of specific expressions into a type of “person name: place name” and sets the classified specific expressions. The type of output.

  When the classifier 25 acquires the processing result of the relationship estimation information acquisition unit 23 and the processing result of the model selection unit 24, a plurality of models are obtained based on the type of the unique expression set selected by the model selection unit 24. A model is extracted from the stored model storage unit 40. Then, the classifier 25 determines, using the extracted model, whether or not each relation information candidate becomes relation information of a set of unique expressions (step S10 in FIG. 3).

  Here, the model uses the analysis processing unit 10 and the result including the result determined in advance for the relationship information corresponding to the predetermined specific expression set and the text including each specific expression corresponding to the predetermined specific expression set, and It is generated in advance by performing well-known machine learning using information extracted in advance from the relationship information extraction process 20. In addition, the discrimination result for a predetermined set of specific expressions is made in advance based on human judgment. Note that each model may be configured according to the type of set of unique expressions, such as “person name: place name”, “person name: name”, or the like, or may be configured without distinguishing the types.

  In this case, since the relationship estimation information given by the relationship estimation information acquisition unit 23 is also used for discrimination by the classifier 25, each relationship information candidate is a set of unique expressions based on the degree of estimation as the relationship information. It can be determined whether or not the relationship information is obtained. In the present embodiment, it is determined that the relationship information candidate “speech” among the relationship information candidates is the relationship information of the specific expression group “Junichiro Koizumi: Tokyo Station”.

  Note that the classifier 25 may be configured to output a numerical value indicating the degree of possible relation information in addition to the determination result of whether or not the relation information is obtained. As machine learning, a well-known machine can be used. However, it is desirable to use a machine that can learn by directly inputting data of a tree structure or a graph structure.

  The relation information extraction unit 26 extracts relation information candidates determined by the classifier 25 as relation information among the relation information candidates as the relation information of the set of specific expressions, and uses the extracted relation information as a display unit. Output (step S11 in FIG. 3). When the classifier 25 is configured to output a numerical value indicating the degree of possible relation information, the related information extraction unit 26 extracts a related information candidate having the largest numerical value as related information. Alternatively, relationship information candidates whose numerical values are larger than a predetermined threshold value may be extracted as relationship information.

  In this way, if the text “Junichiro Koizumi delivered a speech at Tokyo Station and Ichiro Ozawa delivered a speech at Osaka Station” and a set of specific expressions “Junichiro Koizumi: Tokyo Station” were entered, it would be called “Speech”. Relationship information is extracted.

  The present inventor performed performance evaluation of the related information extraction technique described in Japanese Patent Application Laid-Open No. 2008-225666, and found that the optimal relation information extraction accuracy was 19% and the related information was reproduced in the same case. An evaluation result of 33% was obtained. On the other hand, when the performance evaluation is performed for the present invention, the optimum accuracy of extracting the related information is greatly improved to 68%, and the reproducibility of the related information in the same case is improved to 41%. was gotten.

  As described above, in the above-described embodiment, at least one independent word included in the input text is extracted as the relationship information candidate, and the relationship estimation information and the analysis result indicating the degree that the relationship information candidate is estimated to be the relationship information. Since the relationship information is extracted from the relationship information candidates based on the relationship information, the relationship information between the unique expressions can be extracted from the input text. Thus, for example, in the case of “Junichiro Koizumi gave a speech at Tokyo Station and Ichiro Ozawa at Osaka Station”, information representing a temporary relationship “speech” to the specific expression “Junichiro Koizumi” and “Tokyo Station” Can be extracted, and the relationship information between the unique expressions according to individual cases can be extracted.

  In addition, according to the present invention, the extraction accuracy of the optimum relationship information is greatly improved from 19% to 68% as compared with the technique described in Japanese Patent Application Laid-Open No. 2008-225566, and the relationship information in the same case is improved. A special effect was obtained in that the recall was improved from 33% to 41%.

  Furthermore, a relationship estimation information storage unit 30 that stores relationship estimation information corresponding to the relationship information candidate in advance is provided, and the relationship information extraction processing unit 20 stores the relationship estimation information corresponding to the extracted relationship information candidate. Since it acquires from the part 30, relationship estimation information can be acquired easily and the processing efficiency at the time of extracting relationship information can be improved.

  Furthermore, the relational information extraction processing unit 20, when each unique expression is included in the same phrase, includes an independent word adjacent to the front of one of the specific expressions that appears first in the input text, The independent word that is adjacent to the other of the unique expressions and the independent words that are adjacent to each other are extracted as related information candidates, so that a plurality of related information candidates can be extracted from the same phrase including each specific expression. Thus, the extraction accuracy of related information can be improved.

  In addition, when each specific expression is included in different clauses, the relationship information extraction processing unit 20 includes independent words adjacent to the front or rear of one specific expression that appears first in the input text among the specific expressions, Since independent words adjacent to the front or rear of the other specific expression among the specific expressions are extracted as related information candidates, a plurality of related information candidates can be extracted from the clauses including each specific expression. The extraction accuracy can be improved.

  Furthermore, when the main word of the phrase that does not include each unique expression is an independent word, the related information extraction processing unit 20 extracts the morphemes from the head of the phrase to the main word as related information candidates. Can be extracted from clauses that do not include each unique expression, and the extraction accuracy of relation information can be improved.

  In addition, the said embodiment is only a specific example of this invention, and this invention is not limited only to the said embodiment. For example, the present invention installs a program for realizing the functions shown in the configuration diagram of FIG. 2 or a program having the procedure shown in the flow of FIG. 3 through a recording medium or a communication line in a known computer. This is also possible.

  In addition, the related information candidate extracting unit 21 includes a predetermined particle included at the end of one of the consecutive clauses that appears first in the input text, and a predetermined verb included in the other of the consecutive clauses. When included in the beginning of a phrase, a morpheme sequence including one phrase and a morpheme from the beginning of the other phrase to the main word may be extracted as a related information candidate. For example, the relationship information candidate extraction unit 21 stores a functional verb “effect” and a particle “wo” corresponding to the functional verb. When the phrase “influence” and the phrase “influence” appear consecutively in the input text, the relationship information candidate extraction unit 21 selects one of the phrases “influence” and the other “influence”. The morpheme from the beginning of the clause to the main word, that is, the morpheme string “influence” consisting of “influence” is extracted as a candidate of related information.

  In this case, since a morpheme sequence consisting of a combination of a predetermined particle and a functional verb can be extracted as related information candidates, the number of related information candidates can be increased, and the accuracy of extracting related information can be improved. .

  Further, the model selection unit 24 may be configured to select a model type according to the positional relationship of each unique expression. For example, when each specific expression is included in the same phrase, the model selection unit 24 classifies the set of specific expressions into a type of “same phrase type”. Further, the model selection unit 24 classifies the set of specific expressions into a type of “dependency type” when the front specific expression of each specific expression relates to the backward specific expression, and sets the “same phrase type” and “dependency type”. If it does not fall under “type”, the set of proper expressions is classified as “other types”.

  Needless to say, the model selection method described above and the model selection method according to the embodiment may be used in appropriate combination.

The figure which shows the structure of the conventional related information extraction apparatus. 1 is a configuration diagram of a related information extraction device according to an embodiment of the present invention. Relationship information extraction process flow chart Diagram showing the summary of analysis results by the dependency analysis unit The figure which shows the outline | summary of the processing result by a specific expression matching part The figure which shows the outline of the processing result by the dependency structure information acquisition section The figure which shows the outline of the processing result by the dependency structure information acquisition section The figure which shows the outline of the processing result by the dependency structure information acquisition section The figure which shows the outline of the processing result by the relation estimation information acquisition part

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Analysis process part, 11 ... Morphological analysis part, 12 ... Dependency analysis part, 20 ... Relation information extraction process part, 21 ... Relation information candidate extraction part, 23 ... Relation estimation information acquisition part, 26 ... Relation information extraction part, 30: A relationship estimation information storage unit.

Claims (10)

An apparatus for extracting information related to a plurality of input proper expressions,
When a text including each of the specific expressions is input, an analysis processing unit that analyzes morphological analysis of the input text, and analyzes a dependency of a clause constituting the input text,
When the analysis result by the analysis processing unit is acquired, at least one independent word included in the input text is extracted as a related information candidate, and the relationship estimation information indicating the degree to which the related information candidate is estimated to be related information, A relationship information extraction apparatus comprising: a relationship information extraction processing unit that acquires each extracted relationship information candidate and extracts relationship information from the relationship information candidate based on an analysis result and relationship estimation information. A relationship estimation information storage unit that stores in advance relationship estimation information corresponding to the relationship information candidates;
The relationship information extraction apparatus according to claim 1, wherein the relationship information extraction processing unit acquires relationship estimation information corresponding to the extracted relationship information candidate from the relationship estimation information storage unit. The relation information extraction processing unit exists between each unique expression and an independent word adjacent to the front of one of the specific expressions that appears first in the input text when each specific expression is included in the same phrase. The related information extraction apparatus according to claim 1, wherein an independent word adjacent to the rear of the other unique expression is extracted as a related information candidate. The related information extraction processing unit, when each specific expression is included in different clauses, the independent word adjacent to the front or rear of one specific expression that appears first in the input text among the specific expressions, and each specific expression 4. The related information extraction apparatus according to claim 1, wherein a self-supporting word adjacent to the front or rear of the other specific expression is extracted as a related information candidate. 5. The relation information extraction processing unit extracts a morpheme from the head of the clause to the main word as a related information candidate when the main word of the phrase that does not include each unique expression is an independent word. 5. The related information extracting device according to any one of 1 to 4. The related information extraction processing unit includes a predetermined particle at the end of one of the consecutive phrases that appears first in the input text, and a predetermined verb of the other phrase among the consecutive phrases. 6. The morpheme sequence consisting of one clause and a morpheme from the beginning of the other clause to the main word when it is included at the beginning is extracted as a relational information candidate. Relationship information extraction device. A method of extracting information related to a plurality of input unique expressions using a computer,
When the text including each unique expression is input, the computer morphologically analyzes the input text and analyzes the dependency of the clause constituting the input text,
At least one independent word included in the input text is extracted as a related information candidate, and relationship estimation information representing the degree to which the related information candidate is estimated to be related information is acquired for each extracted related information candidate And extracting relation information from the candidate relation information based on the analysis result and the relation estimation information.   The program for functioning a computer as each means of the related information extraction apparatus in any one of Claims 1 thru | or 6.   The program for making a computer perform each process of the relationship information extraction method of Claim 7.   10. A computer-readable recording medium on which the program according to claim 8 or 9 is recorded.

Images