摘要 |
<P>PROBLEM TO BE SOLVED: To improve accuracy of a task for classifying a document into any category in a predetermined category set. Ž<P>SOLUTION: When a category set with which a set of documents corresponding to a category for each category is associated is inputted, a document classification method calculates each document vector of the category for each category as a center of balance of word vectors acquired from a word concept base of words in each document, clusters the document vectors acquired by a document vector acquisition means of each document of the category for each category, acquires a sub-category set in which each cluster of the document vectors acquired as the clustering result is set as a sub-category, and acquires a center of balance of the document vectors, of each document in the sub-category, acquired by the document vector acquisition means as a sub-category vector of the sub-category for each sub-category, of the category, acquired by a document clustering means. Ž<P>COPYRIGHT: (C)2010,JPO&INPIT Ž
|