DOCUMENT CLASSIFICATION DEVICE AND PROGRAM

• Main Authors: IWASAKI HIDEKI, TAIRA HIROSHI, GOTO KAZUYUKI, MATSUMOTO SHIGERU, MIYABE YASUNARI
• Format: Patent
• Language: English
• Published: 24.09.2010
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

PROBLEM TO BE SOLVED: To prepare a category structure along the intention of a user, and to re-classify a document belonging to an integrated category by estimating the intention of the user from the integrating operation of the user. SOLUTION: A category integration part 32 integrates a plurality of categories specified by the user into one integrated category. A user intention estimation part 321 calculates the degree of the attention of a word for each word included in category names of the plurality of categories specified by the user on the basis of the number of category names including the word among the category names of the plurality of categories specified by the user. The user intention estimation part 321 estimates the word which is high in the calculated degree of attention as an attention word. A slave category classification part 322 calculates a co-occurrence frequency of the estimated word and the word included in the document belonging to the integrated category. The slave category classification part 322 classifies the document belonging to the integrated category into a slave category positioned below the integrated category on the basis of the calculated co-occurrence frequency. COPYRIGHT: (C)2010,JPO&INPIT