Research On Domain Ontology-based Intelligence Retrieval Technology

Currently, the World Wide Web's information grow explosively every day, how to obtain valuable information efficiently is biggest challenge. Traditional information retrieval system searching information mainly based on content categories and keyword-based. The system based on keywords mainly depends on the user's experienc, but keywords are prone to be semantic confusion. And retrieval system of content-based categories has been joined by artificial human knowledge, but joining the human knowledge is difficult to ensure timeliness and accuracy. Therefore, the traditional information retrieval system is always unsatisfactory.To deal with problems of traditional information retrieval system, this research designs a framework of domain ontology-based information retrieval system. Based on the definition and classification of domain knowledge, the framework constructs domain ontology, uses Oracle 11g to deposit the concept sets and the relationship sets among the concepts of domain ontology and infers the related knowledge with a user-defined rulebase, which improves retrieval accuracy.The research contents are as follows:Deep study the semantic web and the basic theory of ontology technology, such as the basic elements of ontology, ontology language and ontology building tools. Deep study the method of building ontology, and use the protégébuild domain ontology. To construct the domain intelligence search engine and to release full potential energy of the semantic on search engine will need professional search engine better than the large general network information retrieval services. Therefore, it is necessary to establish the full, complete and authoritative domain ontology library and be full use of the library and reflect the characteristics of the field in order to achieve better results. Using custom defining rules, achieving up reasoning using jena, using its logical and semantic analysis tools and platforms can provide the basis for the Semantic Web, make search engine re-design more intelligent, reflect the user's demand better and improve search precision ratio result. Using JSP technology, based on Eclipse3.1 + Tomcat5.0 + Lomboz3.0 building ontology retrieval systems can achieve the interaction of domain users and servers.The research results are as follows:This study not only build the domain ontology knowledge library, but also realize domain search engines. Because of using jena reason machine and using user-defined rules to reason, the system improve the recall and precision rates. The system also uses JSP technology to strengthen the interaction of the domain users and server, and the use of ontology library resources is easy to design the intelligent human-machine interface which is accepted by the users. The study also uses Oracle 11g database to persist ontology, which makes it more convenient for us to maintain the ontology easily.