A USER-ADAPTIVE NEURAL-NETWORK SUPPORTING A RULE-BASED RELEVANCE FEEDBACK

Associative mechanisms, such as those based on the use of thesauri, do cument clustering and relevance feedback, are widely employed in infor mation retrieval systems to enhance their effectiveness. They make it possible to retrieve also the documents not directly indexed by the se arch terms. In this paper, a relevance feedback model is defined, base d on an associative neural network in which concepts meaningful to the user are accumulated at retrieval time by an iterative process. The n etwork can be regarded as a kind of personal thesaurus of the user. A rule-based superstructure is then defined to expand the query evaluati on with the meaningful terms identified in the network. The search ter ms are expanded by taking into account their associations with the mea ningful terms in the network.