Research And Realization Of Personalized Information Retrieval Based On Semantic Web

With the rapid development of Internet technology, www resources increases rapidly, it provides us with convenient and also brings us the problem that it's difficult to obtain the useful information. In a certain extent, search engine help people solve the problem, but it brings us disadvantages of providing us too much information and the information it returns to us is not accurate enough. It can't provide targeted service to users. Therefore, it's more important to research personalized information retrieval technology. Personalized information retrieval requires us to gain and represent user interest information accurately, and match the documents by the interest, so the documents which can meet user's personalized needs are returned.In this thesis, the main research aim is that the Semantic Web is used to represent user interests and document information, then the calculation of semantic similarity based on Semantic Web and the personalized search are completed, so the documents which users really need are returned. On this basis, the personalized information retrieval system based on Semantic Web is realized.Firstly, the related theory of Semantic Web and the acquisition, representation and storage of user interests'information in personalized information retrieval are researched. Secondly, the Semantic Web is used to represent the user interests and document information. It provides a semantic similarity calculation method based on Semantic Web and a calculation method of interest value. On this basis, personalized information filtering and personalized search are completed. All these are applied to PIRBSW personalized information retrieval system. Then a design scheme of the system is given, it includes system structure design, division of each function module, system searching process, etc. At last, Java programming language is used to realize the PIRBSW personalized information retrieval system which contains user interface module, document processing module, user interests processing module and personalized retrieval module.The test result shows that this system can return to users personalized information retrieval results according to the user's interests and it can respond promptly to user's interest updates. Compare to existing search engine, the precision is improved by using personalized retrieval based on Semantic Web.