Ontology-based Personalized Information Retrieval System

The difference of user hiberarchy and the multiformity of user type make up of the various requirement to information. In the face of numeric information resource and multimedia information resource, compare with traditional user of books information, people worry about not so much that can not find information resource as that can not find information that most suitable and optimum to them. But most of the retrieval system uses keywords to search the information for all different users, these system use the same standard interface and the same retrieval methods ,so they design the same search results for all different user .Users can not accurately express themselves and they can not express information on specific areas which they interest and needs, even using advanced search technology we can not supply the user with the accurate retrieval results in a relatively short list information ,in practical more than half of the retrieval results are not related to the user needs. One of the main reasons is that: the same keywords for different people, in different areas and different time and place may express different meaning, and therefore one can not accurately reflect the key users of a specific area of the specific information needs. So that general information retrieval combines with the specific areas of information processing technology is needed.This paper researches ontology-based personal user modeling technology and its application in the intelligent information retrieval. As a domain model, ontology provides a common comprehension about the domain concepts and the hiberarchy of concepts, at the same time, its utilization also reduces the dependence on the comprehension technology of natural language. In order to solve the flaws in the existing personalization information service, in foundation of discussing each existing personalization information recommendation manners and contraposing the insufficiency of the current information recommendation manners, we introduce the ontology, and propose a personalize information service model which based on ontology. On one hand we join the ontology when constructing user model, which causes comparison of the user model with the information content to be more convenient and quicker. On the other hand, we use the ontology when recommending the personalization information. By doing so, we increase the semantic information, make up for the current information recommendation technology insufficiency in certain degree, and improve the recommendation information's recall and precision.Contraposing the viewpoint proposed in this paper, we construct the ontology in computer domain, and applies it in the computer domain periodical paper retrieval. The paper content saves by the XML form, in order to fully display the configurative and the semantic function of the XML. The experiment indicates that the personalized information service system joined with the ontology effectively raises the level of personalized information service.