Enriching Web taxonomies through subject categorization of query terms from search engine logs

In this paper, we propose a query-categorization approach to facilitating the engineering process of constructing Web taxonomies. One primary step in taxonomy construction is to acquire the domain-specific terminology terms and the mapping between the subjects and these terms. We introduce a technique for categorizing Web query terms from the logs of on-line search services into a predefined subject taxonomy based on their supposed popular search interests. The obtained experimental results show our technique's effectiveness in reducing the workload of human indexers in constructing Web taxonomies and also show its usefulness in various Web information retrieval applications. (C) 2002 Elsevier Science B.V. All rights reserved.