Query Expansion by Taxonomy

The use of taxonomies and ontologies as a foundation for enhancing textual information base access has recently gained increased attention in the field of information retrieval. The objective is to provide a domain model of an application domain where key concepts are organized and related. If queries and information base objects can be mapped to this, then the ontology may provide a valuable basis for a means of query evaluation that matches conceptual content rather than just strings, words, and numbers. This chapter presents an overview of the use of taxonomies and ontologies in querying with a special emphasis on similarity derived from the ontology. The notion of ontology is briefly introduced and similarity is surveyed. The former can be considered a generalization of taxonomy, while the latter can be seen as an interpretation where aspects of formal reasoning are ignored and replaced by measures reflecting how close concepts are connected, thereby significantly enhancing performance. In turn, similarity measures can be used in conceptual querying. Queries can be expanded with similar concepts, thereby causing query evaluation to be based on concepts from the domain model rather than on words in the query.