Improved Relaxation-based Ontology Matching Negotiation

Abstract

Agent-based Ontology Alignment Negotiation processes aims to generate an alignment through the interaction of two or more agents. When these agents exploit different matching services they can reach incompatible alignments, giving rise to conflicts hence the need to engage in a negotiation to achieve consensus. This consensus is sounder as more conflicts are correctly solved. In this paper, a Relaxation-based approach for Agent-based Ontology Matching Negotiation is explored and compared with the MbA/FDO argument-based approach. The limitations of the original Relaxation approach are addressed and modifications are proposed that increase the circumstances under which conflict resolution may occur, thus generating sounder alignments.