TEXT RETRIEVAL PROGRAM AND SERVER AND METHOD USING NON-RETRIEVAL KEYWORD DICTIONARY FOR RETRIEVAL KEYWORD DICTIONARY

摘要

PROBLEM TO BE SOLVED: To provide a text classification program and server and method which suppresses text information which is not illegal nor harmful from being classified into any illegal and harmful category as much as possible when determining whether text information belongs to an illegal and harmful category according to a preliminarily registered keywords.;SOLUTION: There is provided a text retrieval program for making a computer function as a learning text accumulation means for storing a plurality of valid learning text information which does not belong to a specific category and a plurality of invalid learning text information which belongs to the specific category, wherein the learning text information including retrieval keywords is retrieved, and dependency keywords corresponding to the retrieval keywords are extracted, and the validity ratio of the number of valid learning text information to the number of all the learning text information is calculated for each dependency keyword, and a non-retrieval keyword dictionary for registering the dependency keywords whose validity ratio is equal to or more than a prescribed threshold as non-retrieval keywords is generated. Thus, text information including the non-retrieval keywords as the dependency keywords corresponding to the retrieval keywords is prevented from being retrieved.;COPYRIGHT: (C)2011,JPO&INPIT