Semantic Matching for the Medical Domain

This paper proposes some modifications to the SMatch algorithm that enables the semantic matching of medical terminologies using the Unified Medical Language System (UMLS) as a source of background knowledge. Semantic Matching is the process of discovering set theoretic relationships between differing data elements. Initial results from the domain of anatomy are presented that illustrate how semantic relationships can provide greater information during the ontology alignment process than equivalence relationships alone. The paper concludes by demonstrating how this is beneficial in the medical domain.