Research On Personalized Meta Search Engine Based-on Ontology

The capacity of information has been increasing massively with the development of Internet. The users wants to acquire a more overall and accurate result, have to call some search engines again and again. To obtain resources accurately, it requires the search engine to collect the user's hobby and provide his most interested information to the user, thus achieves the goal of provide personalized service to users. The meta-search engineer guarantees the search engineer's precision ratio because users can recall numbers of search engineers according to the meta-search engineer. But many meta-search engineers lacks of personalized service, users may get great numbers of useless information, the search service also has low precision ratio.In this paper, we tried to design a personalized meta-search engineer to improve problem of the traditional search engineer's lack of personalize service. The main work in this thesis can be described as follows:(1)Based on the analysis on developing situation of meta-search engine, this paper addresses some of the existing problems in recall ratio, precision ratio and ignorance of users' interest.(2) In this thesis, we particularly introduces ontology and the knowledge of user interest model and gives the methods to construct user interest model. Design and realize the user interest model, and give the algorithm of calculate user's interest.(3) In the aspects of merging the searching results, mostly existing meta search engine dose not consider user interest, in this thesis, we sorts the searching results from the member search engines based on relevance of user interest, we puts forward the Abstract/Position algorithm based on User Model(APUM).(4) Based on the former work, we build a personalization meta search engineer based on User Model. The experiment results shows that our user model can accurately describe user's interests, our search engine can bring results to users at a higher recall and precision ratio.