What four million mappings can tell you about two hundred ontologies

Abstract. The field of biomedicine has embraced the Semantic Web probably more than any other field. As a result, there is a large number of biomedical on-tologies covering overlapping areas of the field. We have developed BioPortal— an open community-based repository of biomedical ontologies. We analyzed on-tologies and terminologies in BioPortal and the Unified Medical Language Sys-tem (UMLS), creating more than 4 million mappings between concepts in these ontologies and terminologies based on the lexical similarity of concept names and synonyms. We then analyzed the mappings and what they tell us about the ontologies themselves, the structure of the ontology repository, and the ways in which the mappings can help in the process of ontology design and evaluation. For example, we can use the mappings to guide users who are new to a field to the most pertinent ontologies in that field, to identify areas of the domain that are not covered sufficiently by the ontologies in the repository, and to identify which ontologies will serve well as background knowledge in domain-specific tools. While we used a specific (but large) ontology repository for the study, we believe that the lessons we learned about the value of a large-scale set of mappings to ontology users and developers are general and apply in many other domains. 1 Why Create the Mappings? The field of biomedicine has embraced the Semantic Web probably more than any other field. Ontologies in biomedicine facilitate information integration, data exchange, search and query of heterogeneous biomedical data, and other critical knowledge-intensive tasks [12]. As a result, there is a large number of biomedical ontologies covering over-lapping areas of the field [3]. Creating mappings among ontologies by identifying sim-ilar concepts is a critical step in integrating data and applications that use different ontologies. With these mappings, for example, we can link resources annotated with terms in one ontology to resources annotated with related terms in another ontology, discovering new relations among the resources themselves (e.g., linking drugs and dis-eases)