Deep Web Interface Integration Study Based On Domain Ontology

With the rapid development of information technology, the amount of accumulated data is rapidly increasing,how to effectively make use of the information in the Deep Web has become a hot issue of current research. Deep Web data integration is an important measure to make use of the Deep Web information efficiently. Interface Integration as an important part of the Deep Web data plays a unique"nexus"role. So far, Interface integration still has some problems and deficiencies, such as the method form is single, which over-relies on the semantic information of Interface or does not apply the semantic information. In the single field, the Chinese Semantic calculation is not accurate enough and the interface pattern matching method is complex, time-consuming, and so on. Based on the above analysis of shortcomings and deficiencies, this paper presents an approach of Interface Integration based on the domain ontology. This method contains three parts that the matching of the keywords of interface tag, the detection of based on the case and the multi-condition combination query testing.In this paper, a simple and intuitive relationship by OWL language described model is used to build the domain ontology at first with 573 classical instances. And then the Jena toolkit is used to read the domain ontology having been built. What's more, in order to make the examples of ontology having ordered, we take advantage of the Racer reasoning engine embed in the Protégéto achieve the bulk of reasoning in the process of building domain ontology. On the other hand, according to the development needing we have built a number of reasoning rules, and then use the Jena reasoning engine to invoke these rules to reason. These reasoning mechanisms are applied to the Book Search reasoning to reduce the book ontology relations within the connection, greatly reduce the system's response time and enhance the overall integrated system of real-time response speed. We predefine some synonymous about five concept class during the process of building domain ontology, and then make use of these synonyms for the strings matching to complete part of the interface integrate. In addition, for those who cannot keyword matching the label name of clustering, we adopt the detection method based on the cases for interface integration. At last, the way of the combination query is utilized to test the integrated interface. Experimental results show that the method of this paper improves the problems and deficiencies of the current interface integration, has better integrated results and can better meet the user's application requirements.