| 摘要 |
<P>PROBLEM TO BE SOLVED: To determine errors of usage of a postpositional particle in a Japanese-language document. <P>SOLUTION: A Japanese language processing part 2 performs modification analysis of document data input by a document input part 1. A postpositional particle extraction part 3 extracts modification information composed of sets of particles, verbs corresponding to the case particles, and nouns relative to verbs through the case particles, from the analysis result. A case information output part 4 retrieves a list of particles capable of being connected to verbs of extracted modification information, from a case information connection dictionary in a case information connection dictionary storage part 5. A correctness/error determination part 5 reads a discrimination rule from a decision rule storage part 7 and collates the discrimination rule, the list of particles capable of being connected to verbs of extracted modification parts, and the particles connected to verbs corresponding to the list of particles in the extracted modification parts, with one another to decide the usage of particles in respective sentences of input document data to be correct or erroneous and outputs an alarm message showing errors of usage of postpositional particles when erroneous use of postpositional particles occurs. <P>COPYRIGHT: (C)2008,JPO&INPIT |
