A Binary Software Associative Memory

In this paper we describe the implementation of a binary inverted file that acts as a fast software associative memory when combined with a set of efficient search algorithms. We implemented a very fast and powerful retrieval engine, that takes the inverted file scheme to performance levels beyond those possible with the traditional technique. We present the results of testing a production system implementation of this software by the Queensland Police Department. Our results show that the software outperforms conventional techniques in the form of an IDMS implementation, and a hardware solution, namely ICL''s CAFS search engine, in providing immediate response to partial match and complex boolean queries, including text searching. The solution presented is for static files, although the technique may be used with dynamic files too.