Model Research On Intelligent Search Engine Based Ontology

Today, the increasingly expanding of information in Internet makes search engine become an absolutely necessary tool of getting information for people. But, with the increase of information and the manifold of behavior manner, some problems emerged from behind the traditional search modes, such as the problem of glossary isolation, different expressing, faithful expressing, and mechanical matching. Essentially, traditional search engine cannot analyze and extend the meaning of keys user inputs and is short of power to handle knowledge and to comprehend. This paper well uses the theories and technologies of semantic web and ontology, and applies the ability of ontology to establish semantic model to the research of intelligent search. The research of this paper includes the method of data integration based ontology, the model of user interface based concepts, the courses and algorithms of query process, et al. Finally, we present a intelligent search model based ontology SMBDI. Concretely, our work includes several points as follows.Based analysis of associated relationship between information agent DTD and ontology, an ontology data integration method is presented. This method has a goal of gathering all XML information agents to implement a large-scale search and use the path mapping theory to realize semantic integration. It avoids the problem of information losing caused by different expressing, at the same time, using real concepts to identify data also avoids the problem of mechanical matching and promises the accuracy of answer.After considering semantic unification and analyzing user search manner, we designs a user interface based concepts. This UI breaks through the traditional search mode and uses graphic ontology navigation, naturally restriction mechanism and independent output customization. It solves the problem of faithful expressing and makes the communication of human-machine more intelligible.A query split method and it's related query plan build algorithm are presented. For the reason that only part of information covered by a single agent, united query based several information agents is inevitable. Basic idea of our method is to separate child query formoriginal query and use key to keep the relationship between them. This split method solves the problem of information losing essentially and the query plan build algorithm based it also realize query mapping and splitting on the whole information gather, promises the distributing and timely response of query.An application framework based SMBDI is proposed. This framework modularizes the methods presented in SMBDI and arranges it to be several levels. It follows the architecture of semantic web application and offers a reference for the scholars of semantic web application in the field of knowledge share, information retrieval and ontology application.Finally, we implements experiment to test and analysis the key algorithm of query process Qplan and it's optimized algorithm Qplanopt. The experiment proves that out thoughts are correct and feasible.