Assessing thesaurus-based query expansion using the UMLS Metathesaurus

Abstract

Objectives: Assess query expansion using thesaurus relationships and definitions in the UMLS Metathesaurus for improving searching performance.

Methods: The queries from a MEDLINE test collection (OHSUMED) were expanded using synonym, hierarchical, and related term information as well as term definitions from the UMLS Metathesaurus. Documents were retrieved from a word-statistical retrieval system and assessed for recall and precision based on relevance judgments from the test collection.

Results: All types of query expansion degraded aggregate retrieval performance as measured by recall and precision, although 38.6% of the queries with synonym expansion and up to 29.7% of the queries with hierarchical expansion showed improvement.

Conclusions: Thesaurus-based query expansion causes a decline in retrieval performance generally but improves it in specific instances. Further research must focus on identifying instances where performance improves and how it can be exploited by real users.