PURPOSE: To register related information without considering the difference of a case by providing a means to extract a related word pair out of an input character string and judge whether the shift of the case is required or not at the time of the registration of a related word.
CONSTITUTION: The object character string of registration processing is inputted from an input device 101, and is read in a sentence analyzing device 103 through an input information controller 102. After morpheme analysis processing and syntax analysis are executed for the input character string, the extraction of a word pair by a word pair extracting part 106 is executed. In the case that there is the word pair to be registered, it is judged by a shift necessity judging part 107 whether the case of the related particle of that word pair is the shifted object or not. For instance, in the case of 'INU GA KAWARERU' (A dog is kept.), 'GA' is the shifted object, and the related pair of 'INU' and 'KAU' is shifted to the case of 'O'. Then, since 'INU O KAU' and 'INU GA KAWARERU' mean the same contents, the related information extracted first is registered. Accordingly, the related information can be registered without being conscious of the difference of the case.
JPH01131956 | KANA/KANJI CONVERTING DEVICE |
JP2001101175 | METHOD FOR RETRIEVING KANJI |
JPH01114976 | DICTIONARY STRUCTURE FOR DOCUMENT PROCESSOR |