Optimizing Ontology Alignment through Linkage Learning on Entity Correspondences

Data heterogeneity is the obstacle for resource sharing on Semantic Web (SW), and ontology regarded as a solution to this problem. However, since different ontologies are constructed maintained independently, there also exists problem between ontologies. Ontology matching able identify semantic correspondences of entities in ontologies, which an effective method address Due huge memory consumption long runtime, performance existing techniques requires further improvement. In work, extended compact genetic algorithm-based entity technique (ECGA-OEM) proposed, uses both encoding mechanism linkage learning approach match efficiently. Compact does not need store maintain whole population during evolving process, utilization protects chromosome’s building blocks, reduce algorithm’s running time ensure alignment’s quality. experiment, ECGA-OEM compared with participants alignment evaluation initiative (OAEI) state-of-the-art techniques, experimental results show that efficient.