DOCUMENT SEARCH DEVICE AND DOCUMENT SEARCH PROGRAM

摘要

PROBLEM TO BE SOLVED: To provide a document search device and a document search program for comparing documents before normalization and after normalization and predicting normalization by reflecting rewrite tendency of existing normalization.

SOLUTION: A document feature extraction part 103 generates a feature vector of an input document to be normalized 10. A first document similarity calculation part 106 determines the degree of similarity with the feature vector of the document before normalization. A document normalization predicting part 109 weights and averages the feature moving vectors of the documents before and after normalization using the degree of similarity, and determines the feature vector by predicting the normalized document 10. A second document similarity calculation part 110 determines the degree of similarity between the feature vector of the predicted normalized document 10 and the feature vector of the normalized document. A normalization recommendation data generation part 111 recommends the normalized document with a high degree of similarity.

COPYRIGHT: (C)2011,JPO&INPIT