DOCUMENT GROUP PROCESSOR, DOCUMENT GROUP PROCESSING METHOD AND DOCUMENT GROUP PROCESSING PROGRAM

PROBLEM TO BE SOLVED: To provide a document group processor, a document group processing method and a document group processing program, allowing extraction of a word representing a document group. SOLUTION: In this document group processor, a preprocessing part 11 divides each document of the document group into words, and an importance level impartment part 12 imparts an importance level to each word. A score calculation part 13a calculates a ranking score wherein a value increases/decreases according to the ups/downs of a simultaneous co-occurrence degree with an upper word selected as a word of an upper hierarchy to each word except the upper word. A hierarchy-classified keyword determination part 13b determines a lower word to the upper word by order based on the ranking score calculated to each word group except the upper word by the score calculation part 13a. A hierarchical information generation control part 13c makes lower word determination processing by the hierarchy-classified keyword determination part 13b and ranking score calculation processing by the score calculation part 13a to be repeatedly executed over a prescribed hierarchy to perform control to generate hierarchical information wherein the lower word to the upper word is associated over a plurality of hierarchies. COPYRIGHT: (C)2011,JPO&INPIT