Inverted index generation method and apparatus for generating for the N-gram search , retrieval method and retrieval apparatus using the inverted index , as well as , computer program

摘要

PROBLEM TO BE SOLVED: To provide a generation method or the like of a transposition index appropriate for efficiently narrowing down a document including a designated retrieval character string from among a plurality of documents.SOLUTION: The present invention relates to a transposition index generation method including a character counting step, an order switching step and a generation step. In the character counting step, for each of a plurality of items of document data 18 each comprised of an entry word and a corresponding explanatory sentence, the number of characters in the entry word and the explanatory sentence is counted. In the order switching step, an order of the plurality of items of document data 18 is switched into an order starting from the item including most counted characters. In the generation step, "N-gram (N is a natural number) that is a character string of N characters" is extracted from the plurality of items of document data 18 of which the order has been switched, and for each extracted N-gram, a transposition index is generated while making correspondent appearance positions in the plurality of items of document data 18 of which the order has been switched.