The Design And Implementation Of An Intelligent Agent-Based Personalized Meta Search Engine

As the converage of the results that come from most existing single search engine is very limited and also that most of them is unrelevance. Users have to seek help for many Search engines in order to get the more comprehensive, veracious retrieved information. The Meta Search engine technique has solved this problem in a sense. However the meta search engines that exists nowadays has a deficiency in adaptive itself to the user's interest, select suitable member search engine and mergeing the result which makes them hard to meet the user's need.This paper presents the current studying situation and the trend of development of meta search engine, points its deficencies, analyzes the technical characteristics of intelligent Agent sexpatiates its advanced feasibility. We expatiate the relative technology about the personalized meta search engine based on intelligent Agent from the web mining, Information filtering and some other aspect. We put forward an model of personalized meta search engine based on intelligent Agent, and dwell on all parts and principle of the model. As the query expansion, schedule of the member search engine and results merge are the key points in this model. To make the model with high performance of search and more intelligent we improved some algorithms that are corresponding to them to deal with these problems. Via the cooperation between the Multiple Agent and the building of the knowledge base, the system has been provided with following ability: Set up a user interest model pertinently and to allocate users' queries to their interest domain for the sake of extend query in it. Thus the users' queries can be expressed in a more accurate and clear way. Besides, it also provide with personal interface customize, personal information service customize and personal search engine schedule mechanism to meet the difference user's service and need.We implement a rude system and make a few tests on it. Via test, this system has a good performance and efficiency, and it has achieved a good rate of accuracy and speed of search as a good meta search engine. So the practice proofs that this system model is feasible and availablility and it is also able to learn and optimize by itself.