摘要 |
<P>PROBLEM TO BE SOLVED: To provide a new biographic expression identifying technique making it possible to identify personal names even in the case of biographic expressions whose family names cannot be obtained from a biographic dictionary, while preventing errors in specifying personal names due to persons with the same last name but with different first names. <P>SOLUTION: A document to be processed is separated into word units to obtain separate pieces of word information; based on the pieces of word information and rules for extracting biographic expressions, any biographic expression mentioned in the document to be processed, the number of words therein, and name-related information are obtained. For example, a biographic expression consisting of one word is supposed as an abbreviated biographic expression, and biographic expressions including the abbreviated biographic expression are extracted as candidates for the formal biographic expression. If, for each abbreviated biographic expression, there are no candidates for the formal biographic expression, the abbreviated expression is determined to have no formal expression; if there is one candidate for the formal expression, it is determined to be the formal expression; if there are two candidates, the candidate with the corresponding name-related information is determined to be the formal biographic expression. <P>COPYRIGHT: (C)2003,JPO |