| 摘要 |
<P>PROBLEM TO BE SOLVED: To find out a concord relationship being presumed to be truly correct in text data composed of inputted Japanese natural sentence and eliminate ambiguity in the natural sentence, by using a database storing concord pairs forming "concord" in Japanese sentences. <P>SOLUTION: When two or more concord relationships appear in a Japanese natural sentence, based on inputted text data, making the meaning of the sentence ambiguous, from among the distances between words of call elements and response elements in the natural sentence, a response element having a concord relationship, corresponding to an average distance between words of call elements and response elements stored in a concord pair database DB1 is mechanically selected. <P>COPYRIGHT: (C)2005,JPO&NCIPI |
