Combining Two Strategies For Ontology Mapping

Ontologies are the key to the Semantic Web because they are the carriers of the meanings contained in the Semantic Web (McGuinness, 2002). At the same time, ontology mappings can provide a channel from which several ontologies could be accessed and hence could exchange information. Establishing such mappings has been the focus of a variety of research originating from diverse communities. In this paper, we propose an approach ACAOM (A Composite Approach for Ontology Mapping) for automatic ontology mapping based on the combination of name and instance based strategies. ACAOM uses WordNet to calculate similarities between concepts in two ontologies and also uses instances that include text information to build vectors, and then computes similarities. The two similarity measures are then combined to create the results of mapping. The experimental results and comparisons with related work indicate that ACAOM can find mappings effectively.