A New Instance-Based Approach for Ontology Alignment

Abstract

Due to the increasing number of information sources available on the web and their distribution and heterogeneity, ontology alignment became a very important and inevitable problem to resolve in order to ensure semantic interoperability between these sources. Instance-based ontology alignment represents a very promising technique to find semantic correspondences between entities of different ontologies. In practice, two situations may arise: ontologies that share common instances and those share few or do not share common instances. In this paper, the authors describe a new approach to manage the latter case. This approach exploits the reasoning on ontologies in order to create a corpus of common instances. They have used the Biblio and Finance tests of Benchmark series of the OAEI 2012 evaluation campaign to evaluate the performance of their approach. The results obtained show the good performance of the authors' approach compared to ontology alignment systems and improves significantly the instance-based and reasoning-based methods.