摘要 |
PROBLEM TO BE SOLVED: To acquire words of topical interest regarding a designated place without taking time in processing.SOLUTION: This invention calculates all keywords distribution which indicates at which place all the input keywords appear frequently, calculates individual keyword distribution for each keyword at each place, calculates a distribution difference score which indicates distribution difference for each place from the all keywords distribution and the individual keyword distribution and stores them at a keyword score storage means for each keyword. It refers to the keyword score storage means on the basis of the input keywords, acquires the distribution difference score with respect to the input keywords, and selects and outputs the top N cases of the keywords whose distribution difference scores are high. |