摘要 |
PROBLEM TO BE SOLVED: To realize highly precise inter-document similarity calculation by regarding as important an outline inputted with a document. SOLUTION: Two pairs of documents whose similarity should be calculated and the pair of the outlines are inputted (S1), and the morpheme analysis of the inputted two pairs of documents and outlines is operated (S2), and unnecessary words are removed from the two sets of morpheme-analyzed documents and outlines by referring to an unnecessary-word table (S3). Then, the similarity of the two documents is calculated based on words included in each document by weighting words included in each outline (S4). In this case, it is possible to omit unnecessary-word removal. |