ontoMATCH: A probabilistic architecture for ontology matching

There is a great deal of research on the structural and logical matching of ontologies, and there is also considerable work on matching data instances from heterogeneous schema or ontologies, however, little work exploits the fact that ontologies include both data and structure. In this paper, we aim to present a new ontology matching architecture that tightly integrates both semantic similarity measure and structure matching to achieve better matching of ontologies. Based on the observation that semantic similarity measure is a fundamental operation in ontology matching, we first proposed a semantic similarity measure (simDAG) based on WordNet using both the directed graph theory and the probabilistic theory, experimental results show its obvious advantages. Then, we proposed a compound ontology matching architecture (ontoMATCH) considering properties, elements and relationships in ontologies. At last, we evaluate ontoMATCH on the OAEI benchmark set of tests with encouraging results.