| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To reduce a user's labor on narrowing retrieval results. <P>SOLUTION: A computer retrieves names including inputted characters from a database which stores the names, attribute words relating to the names, and relevance which shows the degrees of the relations between the names and the attribute words, and then outputs the retrieved names as the name candidates. When extracting the attribute words relating to the name candidates from the database, the computer computes degree of independence which represents the degrees of the differences between the extracted attribute words, degree of comprehensiveness which represents a degree of comprehensiveness of the name candidates by combinations of the extracted attribute words, and uniformity which represents a degree of uniformness of the number of the name candidates relating to each attribute word, and computes scores of the combinations of the attribute words on the basis of the degree of independence or the degree of comprehensiveness or the uniformity, and then outputs the combinations of the attribute words in higher-score order to an output part. <P>COPYRIGHT: (C)2009,JPO&INPIT</p> |
