Distributed document search apparatus, distributed document search method, and distributed document search program

摘要

PROBLEM TO BE SOLVED: To provide a distributed document search device capable of reducing calculation load for searching by reducing the times of making reference to inverted index.SOLUTION: With resect to plural inverted indexes TIC-TIF, plural phrase signatures A-F, which record if two words appear as a phrase in documents charged to the inverted indexes, are disposed in a tree shaped structure. Plural proximity signature groups gA-gF, which record if two words appear being separated away from each other by a preset distance in documents charged by the inverted indexes, are disposed in a tree shaped structure. When searching a phrase, an iquiry if the two words appear as a phrase is made to upper phrase signatures A and B, and only when the two words appear, the iquiry is made to the lower signatures. When the two words do not appear, the iquiry is not made to thereby reduce the load for search processing.