| 摘要 |
<P>PROBLEM TO BE SOLVED: To support document data analysis by extracting related words and phrases paying attention to composite words. <P>SOLUTION: Data of one group sentences is input by a document input part 10. A morpheme analysis part 11 takes out data of one sentence as a processing object, and performs morpheme analysis. A syntax analysis part 12 performs syntax analysis based on the morpheme analysis result. The syntax analysis result is registered at a syntax analysis result data storage part 13. The above processing is repeated for all sentences. After this, a noun part extraction part 14 extracts morphemes of nouns and composite words of nouns by prescribed extraction rules, and registers them at a noun part extraction result data storage part 16. Retrieval paying attention to only morphemes and retrieval intended for both morphemes and composite words are performed, and retrieval of words and phrases is performed from various viewpoints. <P>COPYRIGHT: (C)2009,JPO&INPIT |
