Compact Evolutionary Algorithm Based Ontology Meta-matching

Abstract

Since different ontology matchers do not necessarily find the same correct correspondences, usually several competing matchers are applied to the same pair of ontology entities to increase evidence towards a potential match or mismatch. How to select, combine and tune various ontology matchers, so-called ontology meta-matching, is one of the main challenges in ontology matching domain. In recent years, Evolutionary Algorithm (EA) based ontology meta-matching technique has become the state-of-the-art methodology to solve the ontology meta-matching problem, but it suffers from some defects like the slow convergence, premature convergence and the huge memory consumption. To overcome these drawbacks, in this paper, a Compact EA (CEA) based ontology meta-matching technique is proposed, which makes use of a probabilistic representation of the population to perform the optimization process. In particular, we construct an optimal model for the ontology meta-matching problem, propose a problem-specific CEA to optimize the aggregating weights of various matchers, and utilize a Cross Sum Quality Measure (CSQM) to adaptively extract the final alignment. The experimental results show that our approach outperforms other EA based ontology matching techniques and Ontology Alignment Evaluation Initiative (OAEI 2016)’s participants.