Semantic Web Fuzzy Ontology Integration Approach Research

Ontology integration, which has been proved to be an effective method, is proposed because of ontology heterogeneous and ontology reuse problems. The traditional research of the Semantic Web ontology is always precise, while human cognition is often imprecise. Fuzziness, as an aspect of imprecision, can be a good expression of the real world knowledge and relationships. Therefore, this article introduces fuzziness to the ontology, and the traditional precise ontology integration extends to the fuzzy ontology integration.In this paper, a Semantic Web oriented fuzzy ontology model is proposed, which promotes the transformation from accuracy to inaccuracy research direction, and not only gives a more detailed formal description of the model, but also highlights the features of the Semantic Web oriented fuzzy ontology itself. A modularized multi-strategy fuzzy ontology mapping method is proposed based on the proposed semantic web fuzzy ontology integration framework aiming at the model. A multi-mapping strategy fusion repair mapping method is designed based on ontology modular. In this method, fuzzy ontology is merged based on mapping results, and the final tightly integration is realized.In order to solve the diffculty of Semantic Web ontology reuse, the modularized ontology is firstly made, and then the mapping and the integration is implemented. Large fuzzy ontology is modularized according to the principle of high cohesion and low coupling, and each module can be regarded as independent small ontology, which can make ontology reuse simple and convenient; the modularized fuzzy ontology is mapped, which can greatly improve the efficiency of mapping, and the result of mapping is higher accuracy and more reliable than before, because multi-strategy mapping method based on structure,attribute,grammar,instances is adopted; Fuzzy ontology is merged from aspects of concepts, attributes, instances and so on according to mapping results, and the membership degrees among the elements of the fuzzy ontologies are remeasured.Accordingly, a prototype system corresponding to the proposed method, which is named as FOIS (Fuzzy Ontology Integration System) is designed and implemented. To verify the effectivity of the proposed model and the usability of the prototype system, the same domain ontologies supplied by the OAEI2010 are utilized in this paper. The experimental results reveal that the design of the FOIS system is better than those of traditional ontology integration systems in terms of discovering more hidden relationships and improving the efficiency and the accuracy of mapping.