JP2011081626A - Dictionary registering device, document label determination system, and dictionary registration program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect a word that expresses a content characteristically and is related to a specific property, or a combination of words from among words appearing in an electronic document while considering relations to other words appearing in the electronic document. <P>SOLUTION: A dictionary registering device includes: an electronic document storage section 110 for storing a set of correct answer documents related to a specific property and a set of incorrect answer documents not related to the property; an input section 176 to which a first word and a second word, namely dictionary registration candidates for determining a document related to the property are input; a document number calculation section 175 for calculating the number the first to eighth documents at each combination of the first and second words; an information amount reference amount calculation section 177 for calculating first to third information volume reference amounts based on the number of the first to eighth documents; and a registration word selection section 178 for comparing the first to third information volume reference amounts and selecting one having the maximum degree related to the property from among the first word and the second word, and a combination of the first word and the second word as a registration word. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a dictionary registration apparatus, a document label determination system, and a dictionary registration program for determining an optimum word for determining whether the content of text information included in an electronic document corresponds to an arbitrary label.

  Conventionally, for text-based web content such as blogs and electronic documents such as document files generated by word processing software, what kind of property the text information contained in the electronic document has? A document label determination system that classifies electronic documents by assigning a label according to the content of the document and using the label is used. Examples of the label include a label indicating a topic of an electronic document such as sports or economy. Among such labels, when determining whether or not an electronic document subject to label determination corresponds to an arbitrary label, dictionary data in which a plurality of index terms having high relevance are associated with the label is used. It is done. For example, when the label is “Economy”, dictionary data in which words such as “Ministry of Finance” and “Exchange” are associated as index words are stored in advance. The document label determination system detects a word that matches the index word included in the dictionary data from the electronic document to be determined, and determines whether the electronic document corresponds to an arbitrary label depending on the degree of the match. judge.

Patent Document 1 discloses a technique for evaluating words that appear in an electronic document, detecting characteristic words indicating the contents of the electronic document, and summarizing the electronic document. Here, the word importance that is highly reliable is calculated by scoring the word that appears in the electronic document according to the combination of a plurality of words that appear in the electronic document and evaluating the word according to the score. is doing.
Further, Patent Document 2 proposes a technique for calculating a word importance by performing a model test based on an information reference amount. Here, a word having a value greater than 0 calculated by subtracting the score calculated by the dependent model from the score calculated by the independent model is selected as an important word.
Further, Non-Patent Document 1 proposes a technique for selecting only words that are important in determining whether a topic corresponds to a model test based on an information amount criterion.

JP 2005-141428 A JP 2005-284209 A

Kazunori Matsumoto, Kazuo Hashimoto, "Schema Design for Causal Law Mining from Incomplete Database", Discovery Science, Second International Conference, DS '99, Tokyo, Japan, December, 1999, Proceedings. Lecture Notes in Computer Science 1721 Springer, pp.92 -102,1999.

However, dictionary data arbitrarily created by the user may be used as the dictionary data to be referred to when determining whether or not a certain electronic document corresponds to a specific label. In such dictionary data, the optimal index word is not always associated with the label, and the index word is fixed, so that it can be flexibly adapted to changes in the contents of electronic documents that change with time. It is difficult to change the index word.
Therefore, from a plurality of electronic documents determined to correspond to a specific label, the words included in the electronic document are analyzed afterwards, and the contents of the dictionary data are recursively according to the words included in the electronic document. A method of updating is conceivable. For example, whether candidate words are scored according to the ratio of the occurrence of any word that is a candidate for an index word in an electronic document and the ratio to other words, and are adopted as index words according to the score. It is possible to decide whether or not. However, although this method scores according to the number of times words appearing in the electronic document appear independently of each other, the words in the electronic document are associated with or related to other words. The semantic content and importance may vary depending on the nature, and it cannot always be said that an accurate index word can be detected. Here, in order to apply the technique disclosed in Patent Document 1 and improve the accuracy of index word detection, a method of generating dictionary data by scoring the word by a combination of a plurality of words is also conceivable. However, there is a problem that as the number of index word candidates increases, the number of combinations increases explosively and the amount of calculation increases.

  The technique disclosed in Patent Document 1 uses a method of x2 test for score calculation, but in the analysis by x2 test, it is necessary to adjust parameters according to the quality of data to be analyzed. Further, with the technique disclosed in Patent Document 2, it is not possible to detect words that are important in determining whether or not a topic is relevant. Further, as in Patent Document 1, there is a problem that the number of combinations explosively increases when the importance of a combination of a plurality of words is obtained.

  In addition, for example, when the score of one word and the score of a combination of a plurality of words are calculated by the technique shown in Non-Patent Document 1 described above, each score has a different scale and depends on one word. The accuracy of the index word and the index word based on a combination of a plurality of words cannot be compared.

  The present invention has been made in view of such a situation. Among words appearing in an electronic document, the contents of the word are characteristically expressed, and a word or a combination of words related to a specific property is expressed in the electronic document. Provided are a dictionary registration device, a document label determination system, and a dictionary registration program that can be detected in consideration of the relationship with other words appearing therein.

  In order to solve the above problems, the dictionary registration device according to the present invention includes an electronic document storage unit that stores a set of correct documents related to a specific property and a set of incorrect documents that are not related to the property. , An input unit for inputting a first word and a second word that are candidates for dictionary registration for determining a document related to the property, and each combination of the first word and the second word Includes the first document number that is the number of the correct documents including both the first word and the second word, and the first word is included and the second word is not included. A second document number that is the number of correct documents, a third document number that is the number of correct documents that do not include the first word and include the second word, and the first word, A fourth document number that is the number of the correct documents not including any of the second words, the first word, and the second word; And a sixth document number that is the number of incorrect documents that include the first word but does not include the second word. And a seventh document number that is the number of incorrect documents that do not include the first word and include the second word, and includes both the first word and the second word. A document number calculation unit that calculates an eighth document number that is the number of incorrect documents that are not valid, and the first to eighth document numbers calculated by the document number calculation unit, A first information amount reference amount indicating a degree of relationship between one word and the property; a second information amount reference amount indicating a degree of relationship between the second word and the property; and the first information amount Information amount reference amount calculation for calculating a third information amount reference amount indicating the degree of relationship between the word and the combination of the second word and the property And the information amount reference amount calculated by the information amount reference amount calculation unit, and the combination of the first word, the second word, the first word, and the second word And a registered word selection unit that selects, as a registered word, a word having the maximum degree related to the property.

In the dictionary registration device according to the present invention, the first information amount reference amount “AIC (DM1)” indicating the degree of the relationship between the first word and the property is:
MLL = (N11 + N12) log (N11 + N12) + (N13 + N14) log (N13 + N14) + (N21 + N22) log (N21 + N22) + (N23 + N24) log (N23 + N24) + (N11 + N13 + N21 + N24) 2 × ZlogZ,
AIC (DM1) = − 2 × MLL + 2 × 4,
Calculated by the following formula:
However, 10 that is the bottom of the log is omitted, and Z = N11 + N12 + N13 + N14 + N21 + N22 + N23 + N24, where N11 is the first document number, N12 is the second document number, N13 is the third document number, and N14 is The fourth document number, N21 is the fifth document number, N22 is the sixth document number, N23 is the seventh document number, and N24 is the eighth document number. .

In the dictionary registration device according to the present invention, the second information amount reference amount “AIC (DM2)” indicating the degree of the relationship between the second word and the property is:
MLL = (N11 + N13) log (N11 + N13) + (N12 + N14) log (N12 + N14) + (N21 + N23) log (N21 + N23) + (N22 + N24) log (N22 + N24) + (N11 + N12 + N21 + N21 + N14 + N21 + N21 + N14 + N14 + N14 + N24 + N24 + N24 2 × ZlogZ,
AIC (DM2) = − 2 × MLL + 2 × 4,
Calculated by the following formula:
However, 10 that is the bottom of the log is omitted, and Z = N11 + N12 + N13 + N14 + N21 + N22 + N23 + N24, where N11 is the first document number, N12 is the second document number, N13 is the third document number, and N14 is The fourth document number, N21 is the fifth document number, N22 is the sixth document number, N23 is the seventh document number, and N24 is the eighth document number. .

In the dictionary registration device according to the present invention, the third information amount reference amount “AIC (DM12)” indicating the degree of the relationship between the combination of the first word and the second word and the property is:
MLL = N11logN11 + N12logN12 + N13logN13 + N14logN14 + N21logN21 + N22logN22 + N23logN23 + N24logN24-ZlogZ,
AIC (DM12) = − 2 × MLL + 2 × 7,
Calculated by the following formula:
However, 10 that is the bottom of the log is omitted, and Z = N11 + N12 + N13 + N14 + N21 + N22 + N23 + N24, where N11 is the first document number, N12 is the second document number, N13 is the third document number, and N14 is The fourth document number, N21 is the fifth document number, N22 is the sixth document number, N23 is the seventh document number, and N24 is the eighth document number. .

The dictionary registration apparatus according to the present invention includes a normalization processing unit that performs word normalization processing.
The dictionary registration apparatus according to the present invention includes a morpheme analyzer that extracts words from a document.

  The document label determination system according to the present invention includes any one of the above dictionary registration devices, a dictionary database in which a registered word selected by the dictionary registration device is stored in association with a label representing a specific property, and the dictionary And a label determination device that determines a label corresponding to the input document using a database.

The dictionary registration program according to the present invention determines an electronic document storage unit storing a set of correct documents related to a specific property and a set of incorrect documents not related to the property, and a document related to the property For each combination of the first word and the second word, a computer having an input unit to which a first word and a second word that are dictionary registration candidates for inputting are input. A first document number that is the number of the correct answer documents that include both one word and the second word; and the correct document that includes the first word and does not include the second word. A second document number that is a number, a third document number that is the number of correct documents that do not include the first word and include the second word, the first word, and the second A fourth document number that is the number of correct documents that do not include any of the words, the first word, and the second document And a sixth document that is the number of incorrect documents that includes the first word and that does not include the second word. The number of incorrect documents that are not included in the first word, the second word is included, and the seventh document number that is the number of incorrect documents including the second word, and the first word and the second word. A step of calculating an eighth document number that is the number of incorrect documents that are not included, and the first word and the property based on the calculated first to eighth document numbers A first information amount reference amount indicating the degree of the relationship, a second information amount reference amount indicating the degree of the relationship between the second word and the property, the first word, and the second word Calculating a third information amount reference amount indicating a degree of relationship between the combination of the property and the property, and the calculated information amount base Compare the quantities and register the first word, the second word, and the combination of the first word and the second word that have the highest degree of relation to the property. A computer program for executing the step of selecting words.
As a result, the dictionary registration device described above can be realized using a computer.

  According to the present invention, regarding a combination of a first word and a second word that are candidates for dictionary registration for determining a document related to a specific property, the first word, the second word, A combination of one word and a second word can be evaluated on the same scale, and a word having the maximum degree related to a specific property can be registered in the dictionary database. Thereby, the dictionary database which mixed the word and the combination of the word can be produced | generated. As a result, among the words appearing in the electronic document, the content is characteristically expressed, and the word or combination of words related to a specific property is considered in relation to other words appearing in the electronic document. The special effect of being able to be detected is obtained.

It is a figure which shows the system configuration | structure of the document label determination system by one Embodiment of this invention. It is a figure which shows the concept of the 2 * 2 contingency table which concerns on the embodiment. It is a figure which shows the concept of the 2 * 4 contingency table which concerns on the same embodiment. It is a flowchart which shows the operation example of the document label determination system by the same embodiment.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram showing a system configuration of a document label determination system according to the present embodiment. In FIG. 1, the document label determination system includes a dictionary database 300, a label determination device 200, and a dictionary registration device 100.

  The dictionary database 300 is a storage device that stores dictionary data. The dictionary data is information in which a label such as a defined word is associated with a plurality of index words that are highly relevant to the label. For example, if the label is “economic”, words such as “Ministry of Finance”, “Exchange”, “Exchange rate”, “Trend”, “Dollar”, etc. are associated as index words. The dictionary data may include a plurality of index words associated with respective labels indicating topics such as “politics” and “sports”. Furthermore, it is also possible to include dictionary data in which index words are associated with a topic hierarchized as “soccer”, “baseball”, etc. among “sports”. In addition, the dictionary data may include, for example, a word “harmful” associated with an index word that is considered to be inappropriate for those under 18 years of age. The dictionary data stored in the dictionary database 300 is read when the label determination apparatus 200 performs a label determination process for the electronic document. The dictionary database 300 may be an independent computer device or may be a database application installed in the label determination device 200.

  The label determination device 200 is a computer device that reads dictionary data stored in the dictionary database 300, compares and analyzes the read dictionary data and an input electronic document, and determines a label corresponding to the electronic document. . Here, the input electronic document is, for example, text data such as a text file such as a blog or a document file generated by word processing software. The label determination apparatus 200 receives an input of an electronic document to be a label determination target, and for each label included in dictionary data read from the dictionary database 300, a word that matches an index word corresponding to the label is included in the electronic document. Label determination processing is performed to determine whether or not the electronic document corresponds to an arbitrary label. For example, when the label determination device 200 detects a word corresponding to the label of the dictionary data exceeding a predetermined threshold from text data included in the electronic document to be determined, the label determination device 200 detects the label as the electronic data. Judged as a document label. For example, the label determination apparatus 200 may determine whether one electronic document corresponds to a plurality of labels, and may assign a plurality of labels corresponding to one electronic document.

  The dictionary registration device 100 recursively calculates the optimal index word corresponding to the label based on the electronic document for which the label determination device 200 has performed the label determination and the label, and the dictionary data stored in the dictionary database 300. Is a computer device that updates and stores data. That is, even if the dictionary data stored in the dictionary database 300 is one in which an index word group arbitrarily determined by the user is associated with the label in the initial state, the dictionary data corresponds to the label. Then, by recursively detecting the optimal index word from the determined electronic document and generating dictionary data, for example, even for blog sites and news sites on the web where important words change over time It is possible to generate dictionary data for determining the optimum label according to the change.

  The dictionary registration apparatus 100 includes an electronic document storage unit 110, a normalization processing unit 120, a morpheme analysis unit 130, a morpheme analysis dictionary storage unit 140, a word distribution calculation unit 150, a word distribution table storage unit 160, An index word score calculation unit 170, a document number calculation unit 175, an input unit 176, an information amount reference amount calculation unit 177, a registered word selection unit 178, and a dictionary registration unit 180 are provided.

  The electronic document storage unit 110 stores a set of correct documents related to a specific property and a set of incorrect documents not related to the property. Specifically, the electronic document storage unit 110 corresponds to an electronic document that has been subjected to label determination by the label determination device 200 and a label determination result that indicates whether or not the electronic document has been determined to be a specific label. Attached and memorized. Here, when the electronic document is blog data, the electronic document stored in the electronic document storage unit 110 includes a text body of the blog article, a pictograph, an HTML (HyperText Markup Language) tag, and the like. Image data is not included.

  The normalization processing unit 120 receives the label-determined electronic document stored in the electronic document storage unit 110 and the label determination result as input, and performs normalization processing on sentences and words included in the electronic document. Output electronic documents. The normalization process performed by the normalization processing unit 120 includes, for example, the following processes. First, similar symbols such as a hyphen, a minus symbol, and a long sound symbol are normalized according to a predetermined rule set. Here, normalization is performed by converting all of these into hyphens, for example. Also, all half-width characters are converted to full-width characters. Also, all tab characters are converted to white space characters. Also, the pictograph is converted into a specific character symbol (for example, 0xA2A2). When the electronic document is web data such as a blog article, the HTML tag is removed from the web data. Japanese lower case letters are converted to upper case letters. Here, for example, the lower case “i” is converted to the upper case “i”. However, when words stored in the morphological analysis dictionary storage unit 140 described later are stored in the morpheme analysis dictionary storage unit 140 in a state in which lowercase letters are included, conversion from lowercase letters to uppercase letters is not performed. . In this case, the lowercase letters of the English letters are kept lowercase and are not converted to uppercase.

  The morpheme analysis unit 130 receives the normalized electronic document output from the normalization processing unit 120, the label determination result for the electronic document, and the morpheme analysis dictionary read from the morpheme analysis dictionary storage unit 140. A morphological analysis process is performed on the digitized electronic document, and a document vector table is output. Here, the document vector table is, for example, when the text “My name is Nakamura” is included in an electronic document, and these are morphologically analyzed, and “I”, “No”, “Name”, “Ha” ”,“ Nakamura ”,“ Is ”, etc., the text data is divided into morphemes (the smallest meaningful unit), and each part of speech is determined to associate the part of speech information.

  The word distribution calculation unit 150 generates and outputs a word distribution table based on the document vector table generated by the morpheme analysis processing by the morpheme analysis unit 130. The word distribution table is a word obtained by removing specific parts of speech such as particles and auxiliary verbs from the document vector table in which text data is divided into morphemes by the morphological analysis unit 130 and extracting only words such as nouns to be index words. It is the table | surface by which each frequency of the list | wrist was matched with the frequency which shows the appearance frequency in an electronic document. Here, the word distribution calculation unit 150 performs normalization processing of words extracted from the electronic document. For example, normalization processing of English words is performed, and lowercase letters of English words are converted to uppercase letters. In addition, the katakana word notation fluctuation is converted, and for example, if there is a word such as “tire tire”, it is converted into character data of “diamond”. In this way, by performing normalization processing at the word level after morphological analysis processing, for example, when there is a notation such as “West Japan”, whether this word is one word of “West Japan” or “Western” It is possible to accurately detect whether the word is composed of two words “book”.

Here, the word distribution calculation unit 150 generates one word distribution table for a plurality of electronic documents, and if the normalized character string is not included in the word distribution table, the normalized word Is newly added to the word distribution table. In the word distribution table generated by the word distribution calculation unit 150, the frequency of appearance frequency associated with each normalized word is calculated according to the number of times a specific word appears in the same electronic document. (Tf: term frequency) and a method for calculating the frequency (df: document) depending on whether a specific word appears in one electronic document regardless of the number of times it appears in the same electronic document. frequency)) can be applied. In the present embodiment, a df method for calculating the frequency based on whether or not a word appears in each of a plurality of electronic documents is applied. The word distribution calculation unit 150 stores the generated word distribution table in the word distribution table storage unit 160.
The word distribution table storage unit 160 stores a word distribution table generated by the word distribution calculation unit 150. As described above, the word distribution table is a data table in which each word that is a candidate for an index word is associated with a frequency indicating the frequency of appearance of the word in the electronic document.

  The index word score calculation unit 170 calculates the score of each word included in the word distribution table based on the label determination result stored in the electronic document storage unit 110 and the word distribution table stored in the word distribution table storage unit 160. Is calculated. The index word score calculation unit 170 includes an SSS calculation unit 171, an MSS calculation unit 172, an MDS calculation unit 173, and an SDS calculation unit 174.

  The SSS calculation unit 171 reads the label determination result stored in the electronic document storage unit 110 and the word distribution table stored in the word distribution table storage unit 160, and whether or not a specific word w is included in the electronic document. A 2 × 2 contingency table is generated for ka, and a single static score word list SSS (W) based on the generated 2 × 2 contingency table is calculated.

FIG. 2 is a diagram illustrating the concept of the 2 × 2 contingency table generated by the SSS calculation unit 171. Here, the number of all electronic documents that are subject to label determination for a specific label (hereinafter referred to as a target label) by the label determination apparatus 200 is N _ALL, and the electronic that is determined to fall under the target label in N _ALL Let N _OK be the number of documents, and N _NG be the number of electronic documents that are determined not to fall under the target label in N _ALL . A set of all the words included in the word distribution table is a word set W, and each word included in the word set W is w. Here, it is assumed that the number of documents including the word w among the N _OK electronic documents determined to correspond to the target label is a. Also, let c be the number of documents that include the word w among N _NG electronic documents that are determined not to fall under the target label. In addition, it is assumed that the number of documents that do not include the word w among the N _OK electronic documents determined to correspond to the target label is b. Also, let d be the number of documents that do not include the word w among N _NG electronic documents determined not to fall under the target label.

At this time, the following equation holds.
A + c = df (w) (the number of documents including the word w in N _ALL )
B + d = N _ALL -df (w)
・ A + b = N _OK
・ A + d = N _NG
In the following description, a + c is represented as q. Moreover, a + b is represented as r. Further, a + c + b + d is represented as z.

Based on the generated 2 × 2 contingency table, the SSS calculation unit 171 obtains the log likelihood value MLL ₁ in the case where it is assumed that there is a causal relationship by the following equation (1), and the information amount reference amount AIC (DM ) Calculate the value. Hereinafter, 10 which is the bottom of the log is omitted.

MLL ₁ = loga + clogc + blogb + dlogd−zlogz
AIC (DM) =-2 × MLL ₁ + 2 × 3
(However, 0log0 = 0)
... (1)

Further, the log likelihood value MLL ₂ when it is assumed that there is no causal relationship is obtained by the following equation (2), and an AIC (IM) value that is an information amount reference amount is calculated.

MLL ₂ = qlogq + rlogr + (z−q) log (z−q) + (z−r) log (z−r) −2zlogz
AIC (IM) = − 2 × MLL ₂ + 2 × 2
(However, 0log0 = 0)
... (2)

  Here, based on the AIC (IM) value and the AIC (DM) value calculated by the above formulas (1) and (2), the word importance E (w) is expressed by the following formula (3) or the following formula: Calculate by (4).

When a / (a + c)> b / (b + d)
E (w) = AIC (IM) −AIC (DM)
... (3)

When a / (a + c) <b / (b + d),
E (w) = AIC (DM) -AIC (IM)
... (4)

Then, after calculating the word importance level E (w) for all the words w included in the word set W, a word list SSS (W) in which the word importance level E (w) values are rearranged in descending order is generated. . At this time, the arrangement of the words w in the word list SSS (W) is w ₁ , w ₂ ,... W _NALL , and the single static score sss (w _i ) for the i-th word w _i is E (w _i ). It becomes. In this way, to generate a sss _{(w i)} the words were arranged in descending order list SSS (W).

The SDS calculation unit 174 includes the word list SSS (W) calculated by the SSS calculation unit 171 and the 2 × 2 contingency table corresponding to each word w _i included in the word set W as inputs, and is included in the word set W. The word list SDS (W) of the single dynamic score sds (w _i ) for each word w _i to be calculated is calculated. Here, SSS (W), which is a word list in which _{single sstatic} scores sss (w _i ) are arranged in descending order by the SSS calculator 171, is represented as a set C (C = {w ₁ , w ₂ ,... W _NALL }). ). Also, let L be a set of words in which the respective words w _i included in the word set W are arranged in descending order according to the value of sds (w _i ). In the initial state, L = {} (empty set).

The SDS calculation unit 174 obtains w _i that maximizes sss (w _i ) from the set C. Then, w _i is removed from C (C = C− {w _i }), and sss (w _i ) is set as temporary sds (w _i ) (sds (w _i ) = sss (w _i )). Here, out of N _OK electronic documents determined to correspond to the target label, the number of documents including the word w _i and another arbitrary word w _j is n ₁₁ ( _ij ). In addition, the number of documents including the word w _i and any other word w _j among N _NG electronic documents determined not corresponding to the target label is n ₁₂ ( _ij ). Also, the number of documents that do not include the word w _i and any other arbitrary word w _j among the N _OK electronic documents determined to correspond to the target label is n ₂₁ ( _ij ). In addition, the number of documents that do not include the word w _i and any other arbitrary word w _j among N _NG electronic documents determined not to correspond to the target label is n ₂₂ ( _ij ). Then, the values a, b, c, and d in the 2 × 2 contingency table generated by the SSS calculation unit 171 are updated as follows for other words w _j in C.

A = a−n ₁₁ ( _ij )
C = c−n ₁₂ ( _ij )
B = b−n ₂₁ ( _ij )
D = dn ₂₂ ( _ij )

Then, the word importance E (w _j ) is calculated from the 2 × 2 contingency table for the word w _{j according} to the above formula (1), the above formula (2), the above formula (3), and the above formula (4). To do. The SDS calculation unit 174 obtains w _j having the largest word importance E (w _j ) among the words included in C, and sets the word importance E (w _j ) as sds (w _j ). Append to L (L = L + {w _j }).
SDS calculator 174 until the set C is an empty set, the process for obtaining _{w i} where sss _{(w i)} is the maximum from the set C, the _{w j} where the value of the most sds _{(w j)} increases The process until it is added to the set L is repeated. Thereby, the word list SDS (W) in which sds (w) is arranged in descending order can be obtained. The word list SDS (W) is a list in which all the words w are arranged in the order of scores in a state where the influence of words higher than the word w is excluded.

The MSS calculation unit 172 receives the SSS (W) calculated by the SSS calculation unit 171 and the 2 × 2 contingency table corresponding to each word w _i included in the word set W, and is included in each word set W. The word list MSS (W) of the multi-static score for the word w _i is calculated. Here, G is a set of arbitrary word combinations ω included in the word set W. In the initial state, G = {} (empty set). A variable indicating the number of combinations ω added to G is u. In the initial state, u = 0.

The MSS calculation unit 172 obtains the word w _i having the largest sss (w _i ) from the word set W. In addition to word _{w i,} obtaining the larger the m words _w 1 to w _m single static score (m is any number that is determined). Then, each 2 × 2 contingency table for the combination ω _j of the word w _i and the word w _j (1 ≦ j ≦ m) is calculated. Here, it is assumed that the number of documents including the combination ω _j of the word w _i and the word w _j among the N _OK electronic documents determined to correspond to the target label is a (ω _j ). Also, the number of documents including the combination ω _j of the word w _i and the word w _j among the N _NG electronic documents determined not to correspond to the target label is c (ω _j ). Also, the number of documents that do not include the combination ω _j of the word w _i and the word w _j among the N _OK electronic documents determined to correspond to the target label is b (ω _j ). Further, it is assumed that the number of documents that do not include the combination ω _j of the word w _i and the word w _j among the N _NG electronic documents determined not to correspond to the target label is d (ω _j ).

At this time, the following equation holds.
A (ω _j ) + c (ω _j ) = df (ω _j ) (the number of documents including the combination ω _j in N _ALL )
B (ω _j ) + d (ω _j ) = N _ALL −df (ω _j )
・ A (ω _j ) + b (ω _j ) = N _OK
・ A (ω _j ) + d (ω _j ) = N _NG

Here, during the process of calculating the 2 × 2 contingency table for the combination ω _j repeatedly for each word w _i , the calculation for the 2 × 2 contingency table for the combination ω _j of w _i and w _j is already performed. If yes, do not recalculate. Thus, if the 2 × 2 contingency table is not calculated redundantly for the same combination, the amount of calculation can be reduced. Then, the importance level E (ω _j ) calculated in accordance with the above formula (1), the above formula (2), the above formula (3), and the above formula (4) is MSS (ω _j ), and the word w _i among the combinations omega _j of word _{w j (1 ≦ j ≦ m} ), determine the most value for MSS (omega _j) is increased omega _j. Here, ω _j is added to the combination set G (G = G + {ω _j }). Also, the variable u is incremented (u = u + 1). Here, among the word _{w k} that is included in the word set W except for the word _{w i,} sss the _{(w k)} is the largest word _{w k,} as a word _{w i} of the importance of the determination target, single static in addition to word _{w i} From the process of obtaining _m words w _{1 to} w m having a large score sss (w _j ), the process of obtaining a combination of words having the largest MSS (ω _j ) value and adding it to the combination set G is repeated. Thereby, a word list MSS (W) in which mss (w) is arranged in descending order of scores can be obtained.

The MDS calculation unit 173 receives the word list MSS (W) calculated by the MSS calculation unit 172 and the 2 × 2 contingency table for the combination ω _j of the word w _i and the word w _j included in W as an input word set The word list MDS (W) of the multi-dynamic score mds (ω _i ) for the combination ω _i of each word w _i and other words included in W is calculated. Here, the MSS (W), which is a word list in which the multistatic scores sss (w _i ) are arranged in descending order by the MSS calculation unit 172, is represented as a set C (C = {ω ₁ , ω ₂ ,... Ω _NALL } ). Also, let L be a set of words arranged in descending order according to the value of mds (ω _i ) for the word combination ω. In the initial state, L = {} (empty set).

MDS calculation unit 173, from the set C, obtaining the omega _i that mds (ω _i) is maximized. Then, ω _i is excluded from C (C = C− {ω _i }), and m ss (ω _i ) is assumed to be temporary mds (ω _i ) (mds (ω _i ) = mss (ω _i )). Here, out of N _OK electronic documents determined to correspond to the target label, the number of documents including a word combination ω _i and another arbitrary word combination ω _j is n ₁₁ ( _ij ). . Further, out of N _NG electronic documents determined not to correspond to the target label, the number of documents including a word combination ω _i and another arbitrary word combination ω _j is n ₁₂ ( _ij ). . In addition, among N _OK electronic documents determined to correspond to the target label, the number of documents that do not include the word combination ω _i and another arbitrary word combination ω _j is n ₂₁ ( _ij ). . In addition, among N _NG electronic documents determined not to correspond to the target label, the number of documents that do not include the word combination ω _i and any other arbitrary word combination ω _j is represented by n ₂₂ ( _ij ). To do. Then, the values a (ω _j ), b (ω _j ), c (ω _j ), d (of the 2 × 2 contingency table generated by the MSS calculation unit 172 for other word combinations ω _j in C. ω _j ) is updated as follows:

A (ω _j ) = a (ω _j ) −n ₁₁ ( _ij )
C (ω _j ) = c (ω _j ) −n ₁₂ ( _ij )
B (ω _j ) = b (ω _j ) −n ₂₁ ( _ij )
D (ω _j ) = d (ω _j ) −n ₂₂ ( _ij )

Then, from the 2 × 2 contingency table for the updated word combination ω _j , the word importance E (ω is followed according to the above formula (1), the above formula (2), the above formula (3), and the above formula (4). _j ) is calculated. The SDS calculation unit 174 obtains w _j that maximizes the value of the word importance level E (ω _j ) among the combinations of words included in C, and calculates the word importance level E (ω _j ) as mds (ω _j ). To the set L (L = L + {ω _j }).

MDS calculation unit 173, until the set C is an empty set, the process of obtaining the omega _i which mss (ω _i) is the maximum from the set C, the most value of mds (omega _j) is increased omega _j The process until it is added to the set L is repeated. Thereby, the word list MDS (W) in which mds (ω) is arranged in descending order can be obtained. This word list MDS (W) is a list in which all word combinations ω are arranged in the order of scores in a state in which the influence of words higher than the word combination ω is excluded.

  The document number calculation unit 175 generates a 2 × 4 contingency table for each combination of the first word w1 and the second word w2, which are dictionary registration candidates for determining a document corresponding to the target label, The first to eighth document numbers N11, N12, N13, N14, N21, N22, N23, and N24 are calculated.

FIG. 3 is a diagram illustrating the concept of the 2 × 4 contingency table generated by the document number calculation unit 175. In FIG. 3, the number of all electronic documents that are subject to label determination for the target label is N _ALL, and the number of electronic documents (correct documents) determined to be applicable to the target label among N _ALL is N _OK. Let N _NG be the number of electronic documents (incorrect answer documents) determined not to fall under the target label in _ALL . At this time, the first to eighth document numbers N11, N12, N13, N14, N21, N22, N23, and N24 are defined as follows.

The first document number N11 is the number of documents including the first word w1 and including the second word w2 among the N _OK correct documents.
The second document number N12 is the number of documents that include the first word w1 and does not include the second word w2 among the N _OK correct documents.
The third document number N13 is the number of documents that do not include the first word w1 and includes the second word w2 out of N _OK correct documents.
The fourth document number N14 is the number of documents that do not include the first word w1 and does not include the second word w2 among the N _OK correct documents.

The fifth document number N21 is the number of documents including the first word w1 and the second word w2 out of N _NG incorrect answer documents.
The sixth document number N22 is the number of documents including the first word w1 and not including the second word w2 out of N _NG incorrect answer documents.
The seventh document number N23 is the number of documents that do not include the first word w1 and includes the second word w2 among the N _NG incorrect answer documents.
The eighth document number N24 is the number of documents that do not include the first word w1 and does not include the second word w2 among the N _NG incorrect answer documents.

At this time, the following relational expression holds.
N12 = “the number a (w1) of documents including the first word w1 out of the N _OK correct documents” −N11
N13 = “the number a (w2) of documents including the second word w2 out of the N _OK correct documents” −N11
N22 = “the number c (w1) of documents including the first word w1 among N _NG incorrect documents” −N21
N23 = “number of documents including the second word w2 out of N _NG incorrect documents c (w2)” − N21
・ Z = N11 + N12 + N13 + N14 + N21 + N22 + N23 + N24
Z is the number of all documents.

  a (w1) is a of the 2 × 2 contingency table related to the first word w1 generated by the index word score calculation unit 170. a (w2) is a of the 2 × 2 contingency table related to the second word w2 generated by the index word score calculation unit 170. c (w1) is c of the 2 × 2 contingency table generated by the index word score calculation unit 170 and related to the first word w1. c (w2) is c of the 2 × 2 contingency table related to the second word w2 generated by the index word score calculation unit 170. Therefore, the number-of-documents calculation unit 175 receives from the index word score calculation unit 170 a, c of the 2 × 2 contingency table relating to the first word w1 and a, c of the 2 × 2 contingency table relating to the second word w2. c is obtained, and the document numbers N12, N13, N22, and N23 are calculated by the above relational expression. Thereby, the amount of calculation can be reduced.

  The input unit 176 inputs the first word w1 and the second word w2 to the document number calculation unit 175. Here, as a method of selecting the first word w1 and the second word w2, the score of each word related to the target label calculated by the index word score calculation unit 170 can be used. For example, a single static score word list SSS (W), a single dynamic score word list SDS (W), a multi static score word list MSS (W), and a multi dynamic score calculated by the index word score calculation unit 170 A predetermined number of important words related to the target label are selected from the word list MDS (W) according to one of the word lists, and the first word w1 and the second word w2 are sequentially selected from the selected word group. Is input to the document number calculation unit 175.

  The information amount reference amount calculation unit 177 uses the first to eighth document numbers N11, N12, N13, N14, N21, N22, N23, and N24 calculated by the document number calculation unit 175 to perform the first to third. Information amount reference amounts AIC (DM1), AIC (DM2), and AIC (DM12) are calculated.

  The first information amount reference amount AIC (DM1) is related to the combination of the first word w1 and the second word w2, which are dictionary registration candidates for determining the document corresponding to the target label, and the first label The degree of the relationship with the word w1 is shown. The first information amount reference amount AIC (DM1) is calculated by the equation (5). In the following formulas (5), (6), and (7), 10 that is the bottom of the log is omitted.

MLL = (N11 + N12) log (N11 + N12) + (N13 + N14) log (N13 + N14) + (N21 + N22) log (N21 + N22) + (N23 + N24) log (N23 + N24) + (N11 + N13 + N21 + N24) 2 x ZlogZ
AIC (DM1) = − 2 × MLL + 2 × 4
... (5)

  The first information amount reference amount AIC (DM1) is the first word related to the combination of the first word w1 and the second word w2 that are dictionary registration candidates for determining the document corresponding to the target label. Only w1 can be used as a measure for dictionary registration.

  The second information amount reference amount AIC (DM2) is related to the combination of the first word w1 and the second word w2 which are dictionary registration candidates for determining a document corresponding to the target label, Indicates the degree of relationship with the word w2. The second information amount reference amount AIC (DM2) is calculated by the equation (6).

MLL = (N11 + N13) log (N11 + N13) + (N12 + N14) log (N12 + N14) + (N21 + N23) log (N21 + N23) + (N22 + N24) log (N22 + N24) + (N11 + N12 + N21 + N21 + N14 + N21 + N21 + N14 + N14 + N14 + N24 + N24 + N24 2 x ZlogZ
AIC (DM2) = − 2 × MLL + 2 × 4
... (6)

  The second information amount reference amount AIC (DM2) is a second word related to a combination of the first word w1 and the second word w2 that are dictionary registration candidates for determining a document corresponding to the target label. Only w2 can be used as a measure for dictionary registration.

  The third information amount reference amount AIC (DM12) is related to the combination of the first word w1 and the second word w2 which are dictionary registration candidates for determining a document corresponding to the target label, The degree of relationship with the combination of the first word w1 and the second word w2 is shown. The third information amount reference amount AIC (DM12) is calculated by the equation (7).

MLL = N11logN11 + N12logN12 + N13logN13 + N14logN14 + N21logN21 + N22logN22 + N23logN23 + N24logN24-ZlogZ
AIC (DM12) = -2 x MLL + 2 x 7
... (7)

  The third information amount reference amount AIC (DM12) is the first word related to the combination of the first word w1 and the second word w2, which are dictionary registration candidates for determining the document corresponding to the target label. Only a combination of w1 and the second word w2 can be used as a measure of whether or not the dictionary should be registered.

  The first to third information amount reference amounts AIC (DM1), AIC (DM2), and AIC (DM12) calculated by the above formulas (5), (6), and (7) are subject to smaller values. Indicates that the degree of relationship with the label is large.

The registered word selection unit 178 corresponds to the target label based on the first to third information amount reference amounts AIC (DM1), AIC (DM2), and AIC (DM12) calculated by the information amount reference amount calculation unit 177. In relation to the combination of the first word w1 and the second word w2 which are dictionary registration candidates for determining the document to be used, only the first word w1 should be registered in the dictionary, or only the second word w2 should be registered in the dictionary. Or whether only the combination of the first word w1 and the second word w2 should be registered in the dictionary. Specifically, the first to third information amount reference amounts AIC (DM1), AIC (DM2), AIC (DM12) are compared, and as a result of this comparison,
When the first information reference amount AIC (DM1) is the minimum value, only the first word w1 is set as a registered word,
When the second information amount reference amount AIC (DM2) is the minimum value, only the second word w2 is set as a registered word,
When the third information amount reference amount AIC (DM12) is the minimum value, only the combination of the first word w1 and the second word w2 is set as a registered word.

  The dictionary registration unit 180 registers the registered word selected by the registered word selection unit 178 in the dictionary database 300 with respect to the target label. Thus, the registered word is stored in the dictionary database 300 in association with the target label.

Next, with reference to FIG. 4, the operation example in which the dictionary registration apparatus 100 according to the present embodiment updates the dictionary data stored in the dictionary database 300 based on the label determination result will be described.
The label determination apparatus 200 reads, for example, a plurality of text-based web contents acquired and stored via the Internet as an electronic document to be determined. Then, the label determination apparatus 200 determines whether or not a certain number or more of words corresponding to the label included in the dictionary data read from the dictionary database 300 are included in the label determination target web content. It is determined whether or not the web content is applicable. The label determination apparatus 200 stores the web content subjected to the label determination process and information indicating the determination result in the electronic document storage unit 110.

  The normalization processing unit 120 reads the web content that has been subjected to label determination and the label determination result stored in the electronic document storage unit 110 (step S1), and performs web content normalization processing (step S2). The morpheme analysis unit 130 performs morpheme analysis processing of the normalized electronic document based on the electronic document whose web content is normalized in step S2 and the morpheme analysis dictionary read from the morpheme analysis dictionary storage unit 140. A document vector table is generated (step S3).

  The word distribution calculation unit 150 generates a word distribution table based on the document vector table generated by the morphological analysis unit 130 in step S3 (step S4). Here, the dictionary registration apparatus 100 will start from step S1 if the electronic document memory | storage part 110 exists in the electronic document memory | storage part 110 with the web content and the determination result which are not the update object of a word distribution table | surface 110 (step S5: YES). The process up to step S4 is repeated.

  When the word distribution calculation unit 150 updates the word distribution table based on all of the web content stored in the electronic document storage unit 110 and the determination result (step S5: NO), the index word score is calculated. The SSS calculation unit 171 of the unit 170 performs the SSS calculation process described above. The SDS calculation unit 174 performs an SDS calculation process based on the SSS (W) calculated by the SSS calculation unit 171 and calculates a word list SDS (W) for which the SDS of each word is obtained (step S7). On the other hand, the MSS calculation unit 172 performs MSS calculation processing based on the SSS (W) calculated by the SSS calculation unit 171 and calculates a word list MSS (W) for which the MSS of each word is obtained (step S8). Then, the MDS calculation unit 173 performs MDS calculation processing based on the MSS (W) calculated by the MSS calculation unit 172, and calculates the word list MDS (W) for which the MDS of each word is obtained (step S9).

  Next, in step S10, first, the document number calculation unit 175 sets the first word w1 and the second word w2, which are dictionary registration candidates for determining a document corresponding to the target label, for each combination of the first word w1 and the second word w2. To calculate the eighth document number N11, N12, N13, N14, N21, N22, N23, N24. Then, the information amount reference amount calculation unit 177 uses the first to eighth document numbers N11, N12, N13, N14, N21, N22, N23, and N24 calculated by the document number calculation unit 175 from the first to the first. Third information amount reference amounts AIC (DM1), AIC (DM2), and AIC (DM12) are calculated.

  Next, in step S11, the registered word selection unit 178 is based on the first to third information amount reference amounts AIC (DM1), AIC (DM2), and AIC (DM12) calculated by the information amount reference amount calculation unit 177. The first word w1, the second word w2, the first word w1, the second word w2, and the combination of the first word w1 and the second word w2, which are dictionary registration candidates for determining the document corresponding to the target label, Among the combinations of the word w1 and the second word w2, the one having the maximum degree related to the target label is selected as the registered word.

  Next, in step S12, the dictionary registration unit 180 causes the dictionary database 300 to store the registered word selected by the registered word selection unit 178 as dictionary data for the target label.

  As described above, according to the present embodiment, the first word w1 with respect to the combination of the first word w1 and the second word w2, which are dictionary registration candidates for determining the document corresponding to the target label, It is possible to evaluate the second word w2 and the combination of the first word w1 and the second word w2 on the same scale, and register the one having the maximum degree related to the target label in the dictionary database 300. Thereby, the dictionary database which mixed the word and the combination of the word can be produced | generated. As a result, the following effects are obtained.

  When using a dictionary database composed only of scored words (uni-grams) when determining the topic of an input document, missed documents corresponding to a specific topic are reduced, but overdetection is increased. On the other hand, in the case of using a dictionary database composed only of scored word combinations (bi-grams), overdetection is reduced, but missing is increased. However, according to the present embodiment, a scored word and a combination of scored words can be mixed and used in the same dictionary database. Therefore, when performing topic determination of an input document, excessive detection and missing are not detected. Both can be reduced.

As mentioned above, although embodiment of this invention was explained in full detail with reference to drawings, the specific structure is not restricted to this embodiment, The design change etc. of the range which does not deviate from the summary of this invention are included.
For example, in the above-described embodiment, the combination of the first word w1 and the second word w2 is a dictionary registration candidate, but the present invention is similarly applied to combinations of three or more words as dictionary registration candidates. It is applicable to.

Also, a program for realizing each step shown in FIG. 4 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed, thereby performing dictionary registration processing. You may go. Here, the “computer system” may include an OS and hardware such as peripheral devices.
Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
“Computer-readable recording medium” refers to a flexible disk, a magneto-optical disk, a ROM, a writable nonvolatile memory such as a flash memory, a portable medium such as a DVD (Digital Versatile Disk), and a built-in computer system. A storage device such as a hard disk.

Further, the “computer-readable recording medium” means a volatile memory (for example, DRAM (Dynamic DRAM) in a computer system that becomes a server or a client when a program is transmitted through a network such as the Internet or a communication line such as a telephone line. Random Access Memory)), etc., which hold programs for a certain period of time.
The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.
The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, and what is called a difference file (difference program) may be sufficient.

DESCRIPTION OF SYMBOLS 100 ... Dictionary registration apparatus, 200 ... Label determination apparatus, 300 ... Dictionary database, 110 ... Electronic document memory | storage part, 120 ... Normalization process part, 130 ... Morphological analysis part, 140 ... Dictionary storage part for morphological analysis, 150 ... Word distribution Calculation unit, 160 ... Word distribution table storage unit, 170 ... Index word score calculation unit, 171 ... SSS calculation unit, 172 ... MSS calculation unit, 173 ... MDS calculation unit, 174 ... SDS calculation unit, 175 ... Document number calculation unit, 176 ... Input unit, 177 ... Information amount reference amount calculation unit, 178 ... Registered word selection unit, 180 ... Dictionary registration unit

Claims (8)

An electronic document storage unit for storing a set of correct documents related to a specific property and a set of incorrect documents not related to the property;
An input unit for inputting a first word and a second word, which are dictionary registration candidates for determining a document related to the property;
For each combination of the first word and the second word, the first document number that is the number of the correct documents including both the first word and the second word; A second document number that is the number of correct documents that include one word and does not include the second word, and a number of correct documents that include the second word but not the first word. A third document number, a fourth document number that is the number of the correct documents that do not include any of the first word and the second word, the first word, and the first word A fifth document number that is the number of incorrect documents that include both of the two words, and a fifth document number that is the number of incorrect documents that include the first word and not include the second word. A document number of 6, a seventh document number that is the number of incorrect documents in which the first word is not included and the second word is included, and the first word and the second word Izu A document number calculating unit for calculating an eighth number of documents of a number of not included said incorrect document,
A first information amount reference amount indicating a degree of a relationship between the first word and the property based on the first to eighth document numbers calculated by the document number calculating unit; A second information amount reference amount indicating the degree of the relationship between the word and the property, and a third information amount reference indicating the degree of the relationship between the combination of the first word and the second word and the property An information amount reference amount calculation unit for calculating the amount,
Comparing the information amount reference amount calculated by the information amount reference amount calculation unit, the first word, the second word, and the combination of the first word and the second word Among them, a registered word selection unit that selects a registered word that has the maximum degree related to the property;
A dictionary registration device comprising: The first information criterion “AIC (DM1)” indicating the degree of the relationship between the first word and the property is:
MLL = (N11 + N12) log (N11 + N12) + (N13 + N14) log (N13 + N14) + (N21 + N22) log (N21 + N22) + (N23 + N24) log (N23 + N24) + (N11 + N13 + N21 + N24) 2 × ZlogZ,
AIC (DM1) = − 2 × MLL + 2 × 4,
Calculated by the following formula:
However, 10 that is the bottom of the log is omitted, and Z = N11 + N12 + N13 + N14 + N21 + N22 + N23 + N24, where N11 is the first document number, N12 is the second document number, N13 is the third document number, and N14 is The fourth document number, N21 is the fifth document number, N22 is the sixth document number, N23 is the seventh document number, and N24 is the eighth document number.
The dictionary registration device according to claim 1. A second information amount reference amount “AIC (DM2)” indicating the degree of the relationship between the second word and the property is:
MLL = (N11 + N13) log (N11 + N13) + (N12 + N14) log (N12 + N14) + (N21 + N23) log (N21 + N23) + (N22 + N24) log (N22 + N24) + (N11 + N12 + N21 + N21 + N14 + N21 + N21 + N14 + N14 + N14 + N24 + N24 + N24 2 × ZlogZ,
AIC (DM2) = − 2 × MLL + 2 × 4,
Calculated by the following formula:
However, 10 that is the bottom of the log is omitted, and Z = N11 + N12 + N13 + N14 + N21 + N22 + N23 + N24, where N11 is the first document number, N12 is the second document number, N13 is the third document number, and N14 is The fourth document number, N21 is the fifth document number, N22 is the sixth document number, N23 is the seventh document number, and N24 is the eighth document number.
The dictionary registration apparatus according to claim 1, wherein the dictionary registration apparatus is a dictionary registration apparatus. The third information amount reference amount “AIC (DM12)” indicating the degree of the relationship between the combination of the first word and the second word and the property is:
MLL = N11logN11 + N12logN12 + N13logN13 + N14logN14 + N21logN21 + N22logN22 + N23logN23 + N24logN24-ZlogZ,
AIC (DM12) = − 2 × MLL + 2 × 7,
Calculated by the following formula:
However, 10 that is the bottom of the log is omitted, and Z = N11 + N12 + N13 + N14 + N21 + N22 + N23 + N24, where N11 is the first document number, N12 is the second document number, N13 is the third document number, and N14 is The fourth document number, N21 is the fifth document number, N22 is the sixth document number, N23 is the seventh document number, and N24 is the eighth document number.
The dictionary registration device according to claim 1, wherein the dictionary registration device is a dictionary registration device.   The dictionary registration apparatus according to claim 1, further comprising a normalization processing unit that performs normalization processing of words.   The dictionary registration apparatus according to claim 1, further comprising a morphological analysis unit that extracts a word from a document. The dictionary registration device according to any one of claims 1 to 6,
A dictionary database in which a registered word selected by the dictionary registration device is stored in association with a label representing a specific property;
A label determination device for determining a label corresponding to an input document using the dictionary database;
A document label determination system comprising: An electronic document storage unit storing a set of correct documents related to a specific property and a set of incorrect documents not related to the property, and a dictionary registration candidate for determining a document related to the property A computer having an input unit for inputting a first word and a second word;
For each combination of the first word and the second word, the first document number that is the number of the correct documents including both the first word and the second word; A second document number that is the number of correct documents that include one word and does not include the second word, and a number of correct documents that include the second word but not the first word. A third document number, a fourth document number that is the number of the correct documents that do not include any of the first word and the second word, the first word, and the first word A fifth document number that is the number of incorrect documents that include both of the two words, and a fifth document number that is the number of incorrect documents that include the first word and not include the second word. A document number of 6, a seventh document number that is the number of incorrect documents in which the first word is not included and the second word is included, and the first word and the second word Izu Calculating a number of documents eighth the number of not included said incorrect document,
Based on the calculated first to eighth document numbers, a first information amount reference amount indicating a degree of relationship between the first word and the property, the second word, and the property And a second information amount reference amount indicating the degree of the relationship between the first word and the combination of the second word and the property. Steps,
Comparing the calculated information amount reference amount, the first word, the second word, and the combination of the first word and the second word are related to the property. Selecting the registered word with the maximum degree,
Dictionary registration program for running

Images