Sufferage Extraction Algorithm of Ontology Mapping Based on Reasoning
Si Chang, Qingjun Zhang, Xin Liu, Бо Лю · 2008
Due to output similarity matrix often involves some error mappings when using multi-strategies for ontology alignment, how to extract the mappings from similarity matrix correctly and effectively will have an important influence on precision. In this paper, a novel approach is proposed, which employs Sufferage algorithm combined with logical reasoning to extract mappings. Firstly, distributed description logical is used to check the semantic consistency of candidate mapping set; secondly, repairing rules are given to repair the probable structure conflicts existed in the mappings; finally, Sufferage algorithm is employed to extract mappings. Comparing with existing algorithms, our experiment results show that this algorithm improves the precision significantly, and thus gets a better mapping result.