Matching heterogeneous ontologies with adaptive evolutionary algorithm

Abstract

An ontology provides a formal description on the domain concepts and their relationships. Due to the subjectivity of ontology engineers, one concept might be expressed in various ways, yielding the so-called ontology heterogeneity problem, and ontology matching is a ground method to address this problem. Ontology matching technique uses the similarity measure to determine the correspondences between two heterogeneous ontology entities. In order to improve the quality of ontology alignment, it is necessary to combine different kinds of similarity measures, and how to optimize the aggregating weights is called the ontology meta-matching problem. Tin this work, a heuristic evaluating metric on the ontology alignment is presented to evaluate the ontology alignment's quality, and a mathematical model on ontology meta-matching problem is constructed. Then, an Adaptive Evolutionary Algorithm (AEA) is proposed to effectively solve this problem. In particular, when the elite solution remains unchanged, AEA adaptively activates three independent exploring strategies, which, respectively use the adaptive selection, crossover and mutation operators based on the population diversity metric. In the experiment, we compare AEA among EA based matching technique and the state-of-the-art ontology matching technique, and the experimental results show its effectiveness.

Keywords:

• Ontology heterogeneity
• ontology meta-matching
• heuristic evaluating metric
• adaptive evolutionary algorithm

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 http://jena.apache.org/

2 http://oaei.ontologymatching.org/

Additional information

Funding

This work is supported by the National Natural Science Foundation of China (Grant Numbers 62172095, 61773415, 61801527 and 61103143), the Natural Science Foundation of Fujian Province (Grant Number 2020J01875).