Query Expansion Based on Topics

Query expansion technology and formal concept analysis (FCA) are two effective methods for improving the query precision in the information retrieval field. In this paper, formal concept analysis which aims at modeling concepts is adopted to expand the user query and a new query expansion approach is proposed. We use the TREC as the contexts and build concept lattices. Then the expansion source is extracted from these concept lattices. For extracting the user interest topics, the terms of the expansion source are matched with the nodes of the ODP directory tree. And then the relevance topics are expanded to the query terms as the expansion results.