DEVICE AND PROGRAM FOR DISCRIMINATING DOCUMENTS WITH THE SAME NAME

摘要

PROBLEM TO BE SOLVED: To obtain a retrieval result only for a specific person desired by a user by separating documents of different persons with the same names when information on the person is retrieved.;SOLUTION: The present invention accumulates a document retrieved based on a person name in a document storage means, reads the document from the document storage means to extract a feature vector, stores the feature vector in a document vector storage means, reads the feature vector from the document vector storage means, clusters the feature vector, generates a cluster group of similar vectors, stores the cluster group in a cluster storage means, acquires a document corresponding to a vector included in the cluster vector from the cluster storage means from the document storage means to generate a cluster of the vector, and outputs the cluster of the document.;COPYRIGHT: (C)2009,JPO&INPIT