Ontology Based Scenarios for Query Expansion Systems-A Case Study Using a Basic Clinical Ontology of Breast Cancer

Abstract

This paper proposes an ontology based query expansion model to improve the precision of keyword search for information retrieval. Scenarios are used as search statistics to enhance keyword mapping and to help expanding the original keyword set by linking those unmapped keywords to other highly used search terms. The purpose of using ontology is two-fold. First, it helps create initial scenarios for a search system, and secondly, it helps infer the keyword-related concepts in the ontology that are finally used to expand the keyword set. The proposed model also finds the most matched scenario to further expand the keywords that can not be mapped to any concept in the ontology hierarchy. A keyword set used by a user usually connote the specific intention of the user in a search. When scenarios change frequently, the use of scenarios can also dynamically bias search results to the latest interests of most users, i.e., the recent hottest topics a domain.