Query Expansion And Retrieval Research Of The Domain Ontology

Accompanied by social progress and development,and the Internet play an increasingly indispensable role of life in people's daily life. A variety of information is expanding quickly. Consequently, how to promptly and accurately select information of meeting the needs of users from the miscellaneous vast amounts of information has become the significant challenge for retrieval.The traditional keyword-based retrieval method accomplish merely primitive matching at the lexical level,and numerous issues concerning the retrieval result is inevitability,such as synonym and polysemy. It is impossible to provide more appropriate results for the user. The ontology technology which can be understood as a computer model not only has the support of semantic information,but it also can be reused and shared. The ontology technology is introduced to the tourist information retrieval system. The method gets the utmost out of semantic associations between information as well as retrieves information based on the level of semantics,and improves the performance of the system.Firstly,this paper concentrates on investigation of a series of ontology theories and the semantic web,then briefly outlines the principles,processes,construction tools and methods of ontology construction. The correlative concept involved to tourist domain,and relationships of data attribute and object attribute are modeled. In this these,the tourism domain ontology is constructed.Secondly,this paper research retrieval extension technology and presents a modified retrieval extension method, which bases on the combination of local-corpus-analysis-based query extension methods and ontology technology. Ontology's knowledge expression of domain information makes that this new method achieves the support of semantic,and optimizes process of local corpus analysis. Under the combination of the ontology and local corpus analysis,the relevance and accuracy of the extension result has improved a lot. The concept similarity is calculated and some of the concepts which are larger than specified threshold are added to the sets of search key. The new concept set is presented. With the Jena framework's support,some uses of inference rules of the tourism domain ontology are customized. Using Jena inference engine,the ontology information is retrieved and the searching results are returned to the users.Finally,on the basis of the above technology research,this paper built a travel information retrieval system based on ontology technology:TIRSOO. According to the modified retrieval method invented by the paper,the retrieval extension and reasoning technology is achieved,and the results is returned. Using an experiment to validate the effectiveness of the method,the results show that TIRSOO effectively improve the intelligence.