Studies On Fuzzy Ontology Reengineering Supported By Databases

The Semantic Web is an improvement and extension of current Web.The goal of the Semantic Web is to help machines to understand the semantic of information,and make it possible for efficient information sharing and machine intelligence collaboration.In order to organize the semantic information in the Semantic Web effectively and realize that machines or devices automatically recognize and process the Web data,ontology is usually adopted as the knowledge representation model.With wide application of ontolgoies,a large number of ontologies are established manually,semi-automatically,or automatically.Under the framework of ontology engineering,transformation of ontologies into database models is one of the important issues of ontology management.This process is called ontology reengineering.In the context of ontology reengineering,transformation of ontologies to conceptual data models serves the reuse and integration of ontologies,while the transformation of ontologies to logical database models serves the persistence of large-scale ontologies.Classical ontologies lack the expressive power of representing imprecise and fuzzy information,which are abundant in real world.Therefore,many researchers have introduced fuzzy sets theory into classical ontologies,and thus proposed fuzzy ontology models.On the other hand,fuzzy database is an important form of representing and processing fuzzy information in real-world applications.Fuzzy database technologies have made numerous achievements in fuzzy data representation and processing after years of research and development,which provide a theoretical basis for reengineering fuzzy ontologies to fuzzy data models.This dissertation,from the perspective of meeting the need of fuzzy ontologies management,realizes the transformation from fuzzy ontologies to fuzzy database models by means of ontology reengineering.Firstly,through investigation of fuzzy ontologies,formal definition of fuzzy OWL 2 ontology is proposed.On this basis,according to the need of reengineering of fuzzy OWL 2 ontologies,the approaches of transforming from fuzzy OWL 2 ontologies to fuzzy conceptual data models and fuzzy logical database models are studied.Among fuzzy conceptual data models,fuzzy EER(Extended Entity-Relationship)model and fuzzy UML class diagram model are selected,as representative target models of tuzzy UWL 2 ontologies reengineenng.Among fuzzy logical database modesl,fuzzy relational database model and fuzzy object-oriented database model are selected as target model of fuzzy OWL 2 ontologies reengineering.The cotents and main contributions of this dissertaion as follows:(1)A fuzzy OWL 2 ontology reengineering approach based on fuzzy EER model is proposed to provide basic technical supports for reusing and integration of fuzzy ontologies based on fuzzy EER model.Firstly,a formal definition of fuzzy EER model is given.Then,a formal approach of transforming fuzzy OWL 2 ontologies and instances to fuzzy EER model is proposed.Finally,through case analysis and theoretical proof,it is indicated that the proposed mapping approach is reasonable and feasible.(2)A fuzzy OWL 2 ontology reengineering approach based on fuzzy UML model is proposed to provide basic technical supports for reusing and integration of fuzzy ontologies based on fuzzy UML model.Firstly,through analyzing the structure of UML class diagram model,a formal definition and semantic interpretation of fuzzy UML class diagram model is proposed based on the fuzzy set and possibility theory.On this basis,a formal approach of transforming fuzzy OWL 2 ontologies and instances into fuzzy UML class diagram model is given.Finally,a transformation example is given for illustrating the proposed approach.The correctness of the transformation approach is proved.(3)An approach of transforming fuzzy OWL 2 ontology to fuzzy relational database is proposed.Firstly,a fuzzy ontology storage structure based on fuzzy relational database is proposed to provide supports for large-scale fuzzy ontology persistent storage based on fuzzy relational database.On this basis,a formal approach of mapping fuzzy OWL 2 ontologies into fuzzy relational databases is proposed.Finally,through case analysis and theoretical proof,it is indicated that the proposed approach is reasonable and feasible.(4)An approach of transforming fuzzy OWL 2 ontology to fuzzy Object-Oriented Database(FOOD)is proposed to provide supports for large-scale fuzzy ontology persistent storage based on FOODs.Firstly,the formal definition of FOOD model is given.On this basis,the transforming rules of from fuzzy OWL 2 ontologies to FOOD model is proposed.Finally,a transformation example is given for illustrating the transformation process.The correctness of the transformation approach is proved.Based on the aforementioned works,a complete theoretical framework for fuzzy ontology reengineering is constructed,which lays a solid theoretical foundation for the semantic interoperability between the areas of Semantic Web and databases,and also provides effective technical supports for the implementation of ontology reengineering.