Interactive Cross-Lingual Ontology Matching
Junfeng Chen,Xingsi Xue,Yikun Huang,Xuewu Zhang
DOI: https://doi.org/10.1109/access.2019.2920881
IF: 3.9
2019-01-01
IEEE Access
Abstract:Recently, with the growing number of ontologies defined in different languages, to bridge the semantic gaps between them, it is necessary to identify the correspondences between their heterogeneous entities, so-called cross-lingual ontology matching. Due to the complexity and the intricacy of the cross-lingual ontology matching, it is essential to get an expert involved in the matching process to guarantee the alignment’s quality. In this paper, we propose an interactive cross-lingual ontology matching technique that makes the user and automatic matcher work together to create high-quality alignments in a reasonable amount of time. In particular, we present a cross-lingual similarity metric to calculate the similarity value of two cross-lingual entities, construct an optimal model for the cross-lingual ontology matching problem, and propose an interactive compact differential evolution (ICDE) algorithm to effectively match the cross-lingual ontologies. The experiment exploits the ontology alignment evaluation initiative (OAEI) multifarm track to test our proposal’s performance. The experimental results show that the ICDE significantly outperforms other EA-based matchers and OAEI’s participants, and the interacting mechanism can significantly improve the alignment’s quality.