Abstract:
The problem of ontology mapping has attracted considerable attention over the last few years, as the usage of ontologies is increasing. In this paper, we revisit the fundamental assumptions that drive the mapping process. Based on real-world use cases, we identify two distinct goals for mapping, which are: (i) ontology development and (ii) facilitating interoperability. Most of current research on ontology mapping has been focused on ontology development and is rooted in the seminal work of McGuinness and Noy in 2000. For example, the well studied problem of ontology merging is an ontology development task. Note that with the increase in the number of information systems that utilize ontologies, facilitating interoperability between these systems is becoming more critical. We compare interoperability to the information integration problem in databases. As a result of this comparison, class matching is emphasized, as opposed to the matching of other entities in an ontology. To the best of our knowledge, this is the first work that distinguishes facilitating interoperability, from ontology development and merging.