AN INFERENTIAL APPROACH TO INFORMATION-RETRIEVAL AND ITS IMPLEMENTATION USING A MANUAL THESAURUS

Most inferential approaches to Information Retrieval (IR) have been in vestigated within the probabilistic framework. Although these approach es allow one to cope with the underlying uncertainty of inference in I R, the strict formalism of probability theory often confines our use o f knowledge to statistical knowledge alone (e.g. connections between t erms based on their co-occurrences). Human-defined knowledge (e.g. man ual thesauri) can only be incorporated with difficulty. In this paper, based on a general idea proposed by van Rijsbergen, we first develop an inferential approach within a fuzzy modal logic framework. Differin g from previous approaches, the logical component is emphasized and co nsidered as the pillar in our approach. In addition, the flexibility o f a fuzzy modal logic framework offers the possibility of incorporatin g human-defined knowledge in the inference process. After defining the model, we describe a method to incorporate a human-defined thesaurus into inference by taking user relevance feedback into consideration, E xperiments on the CACM corpus using a general thesaurus of English, Wo rdnet, indicate a significant improvement in the system's performance.