| 摘要 |
PROBLEM TO BE SOLVED: To easily and promptly sort out a plurality of texts from a plurality of unsorted texts the texts with the same or similar tendency stored in an existing database. SOLUTION: Text mining for a plurality of texts within a sub database (DB) of interest in a main DB 11 is carried out to select a combination C consisting of an important keyword showing the tendency of the plurality of texts and related keywords (step I). In a sorting process (step III), based on the combination C and the result of text mining carried out for the plurality of unsorted texts stored in an unclassified DB 12, the texts with the same or similar tendency as the plurality of texts in the sub DB of interest are sorted out from the plurality of texts. A series of procedures for these processes are carried out as the individual sub DBs (11 to 1-n) classified according to predetermined items within the main DB 11 are sequentially given importance, whereby the plurality of unsorted texts are sorted out into a plurality of sub DBs (12 to 1-n) classified the same way as the predetermined items. COPYRIGHT: (C)2004,JPO
|
