| 摘要 |
PROBLEM TO BE SOLVED: To provide a system capable of efficiently calculating relevance between any kinds of keywords, based on co-occurrence between the keywords. SOLUTION: The system 10 for calculating relevance between keywords includes: a keyword extracting part 14 for extracting the keywords from a plurality of document files; and a relevance calculating part 18 for calculating relevance between one pair of keywords concerning any combinations of the keywords, and storing the relevance in a keyword relevance table DB 26. The relevance calculating part 18 calculates the frequencies of appearance actually appearing keywords in units of document files, calculates the square values of the frequencies of appearance for each keyword, tabulating the square values over the whole document files, calculates the product values of the frequencies of appearance between the pair of keywords in units of document files, tabulates the product values over the whole document files, calculates the square roots of the total sums of square values of the respective keywords, adds both square roots, divides the total sums of product values between the keywords by the sum of square roots, and then, calculates the relevance. COPYRIGHT: (C)2009,JPO&INPIT
|
