| Semantic Web is a new orientation of the Web. It has been a hotspot to people's discussion and study, since the concept of semantic web be put forward by Tim Berners-Lee. Along with the development of the semantic web, the application of ontology is common, and the research in ontology is increasingly becoming a pop direction.Ontology is defined to be a formal, explicit specification of a shared conceptualization. It has strong capability of knowledge expressing, sharing, and reusing .which is also the main target to the ontology'applications. However, in the distributed case, a single ontology can not satisfy the need. It is required to inherit multi-ontology so as to support the sharing and the intercommunion among heterogeneous ontologies. For constructing ontology, researchers propose many criteria. The criterion proposed by T R Gruber is one of the most influential. It includes five regulations: clarity and objectivity, coherence, completeness, maximal extendibility, and minimal commitment. However, because there is no common criteria for constructing ontology, also the constructor and the way modeling the ontology isn't homology at all. All of these would be result in lots of heterogeneous ontologies. So the question of ontological heterogeneity is pop out increasingly.On semantic web, in order to get more information which is needed by other application, or to achieve stronger function by associating with multi-applications, the intercommunication of information among different systems becomes more popular and frequent. But intercommunicate information won't be carried out when using the heterogeneous ontologies in system. Now, in the applications of semantic web, the ontological heterogeneity result in lots of questions of information sharing and intercommunication. For that, settling the problem of heterogeneous ontologies and getting rid of the hindrance among the systems are the key issues to the semantic web's facing.For dealing with the hereinbefore questions, It is needed to establish relation among differ ontologies. Ontology mapping presents itself to find out the semantic relation from heterogeneous ontologies. Through the mapping, we establish mapping among the ontology. Evenly we can reach the universal interface and the same understanding when intercommunicating between the formerly and the new field. Now, there are many approaches in researching the ontology mapping. Presented methods, classified as defining model of ontology, are included four points as follows: (1) methods based on syntax; (2) methods based on instance; (3) methods based on defining; (4) methods based on structure. All the four points refer to different parts of the definitions to the ontologies. When classified as the technology the mapping adopting, the mapping including three points: (1) methods based on regular; (2) methods based on statistics; (3) methods based on machine learning. There must to declare is that each approach always consists of multi-techniques and multi-objects. But there is still much limitations for improvement. such as the faultiness, the excessive compution and the unilateralism of similar calculation.In this paper we proposed an ontology mapping approach SMTOM which combines three strategies. The SMTOM can achieve the mapping of heterogeneous ontologies preliminary when using OWL DL as the logic language. First of all, the paper introduces some concepts and approaches which related to the semantic web and the mapping of ontologies, analyses the new development in the present semantic web, and presents the status of study in the domestic and academic countries. And then contraposes the deficiency of the existing approaches, the paper advances the mechanism of SMTOM and then validated the feasibility and efficiency of the mapping through academic analyses and experiment. We know, for the present semantic web, it's much important to find out the relation, the expression and the performance of the ontology mapping. What's more, the matching itself is an intricate and difficult process. Thereby, this thesis emphasizes particularly on the computation of the semantic similarities and the find of the relation of ontology mapping. Lastly, combining with the service platform which is based on the semantic web, the paper particular illuminates the idiographic application of SMTOM mechanism.The main works in this paper are as follows:(1) Review the background and the status of the ontology mapping. Introduce the technology and the structure related to the semantic web, including the concept and the classes of ontology.(2) Expatiate the concepts and approaches relate to ontology mapping, select the basic approaches that is to be extended to form the algorithms of this paper.(3) Extend the selected approach to form the new approach SMTOM so as to perfect the existing approaches. Then introduce in detail the similarity based on instance, structure and axiom. Evenly we import the experiments to test the feasibility and efficiency of the SMTOM.(4) Apply the extended approach to realize the mapping of heterogeneous ontologies. Present the matching platform and analysis its model functions. Experimentation combining with the tourism service system of Guangxi is demonstrated. The result shows that the extended approach not only resolve the problem which is due to the heterogeneous ontologies but also appeal to the deed of data sharing. Moreover, it conquers the exiting approaches'limitation of unilateralism and improves the information sharing. Evenly elevate the precision and the veracity of information search.(5) Works in this thesis are summarized and future works are pointed out. Furthermore, the research direction and development prospects for ontology and ontology mapping are mentioned.The research on semantic web is a new field. There are several aspects that can be improved in our proposed system. We will try our best to research more logical and efficient mapping approachs to improve the efficiency of the ontological systems. |
PDF Full Text Request