INFORMATION RETRIEVAL DEVICE AND METHOD

摘要

PROBLEM TO BE SOLVED: To update an index file that is used for retrieving on line all sentences and also to reduce the capacity and producing time of the index file by producing the double character chains when the index file of retrieval data is produced for a file of an updated file memory and checking the redundancy of double character chains to delete the redundant double character chains. SOLUTION: An occurrence frequency coefficient unit 703 calculates the occurrence frequencies of registered character strings 1 and 2 for every type of characters and produces the double character chains for both strings 1 and 2 to which the occurrence frequencies are given. A character chain sorter 702 sorts the character chains in every type and aligns the double character chains in each type of characters constructing these character chains in regard to the document number and the occurrence frequencies of the 1st and 2nd characters. A character chain combination calculator 704 calculates a set of double character chains having the redundant document numbers. Then a character chain information store device 706 aligns the number of double character chains, the document number and a set of occurrence frequencies of double character chains and stores them in a character chain information memory 705.