An Improved Algorithm Of Name Strategy And Structure Strategy In Ontology Matching

Ontologies play a crucial role in the emerging Semantic Web, and ontology mapping is a crucial part of ontology query and integration. The thesis focuses on research on these issues.Firstly, the thesis simply introduces the research background of the ontology matching, and summarizes some common ontology matching methods and systems.Secondly, the current name strategies do not consider the semantic ambiguity of word when apply WordNet to calculate similarity between words, and blindness is happened when the current name strategies combine words similarity. To solve this problem, a new solution based on WordNet is proposed in the thesis. The solution maps core words of concept to WordNet to eliminate word ambiguity, and C-C principle is adopted to solve the blindness problem. The solution will improve the accuracy of similarity calculation.Thirdly, the current structure strategies do not consider the different entities that may have different propagation value. To solve this problem, a new algorithm of similarity propagation is proposed, which is based on information content of concept obtained from the structure information of ontology, and it will improve the credibility of structure strategy.Finally, based on the research above, ontology mapping sub-system SNAXMapping2 has been designed and implemented. Experiment result shows that this system improves the precision while maintaining recall.