Study On The Personalized Information Retrieval Technology Based On Intelligent Agent

With the development and the spread of Internet, the information age has come to us. The number of the World Wild Web severs mounts up quickly and doubles every six months. Although the technology of the searching engines has solve the problems such as the information overloading and the resource isotropy etc., and it can finds some useful resource for users, but it's imperfectly to the whole net, the effect can't turn up trumps in many aspects, for example, the covering level of resource, the searching precision, the visualization of searching results and the maintainability etc.The paper firstly introduces the current studying situation of the searching engines,points out its deficiencies, analyzes the technical characteristics of Agent and expatiatesits advanced feasibility and then puts forward a new frame construction of intelligentsearching engines, the system consists of several parts, which are expert Agent,information searching Agent, information filtering Agent, query expanding Agent anduser Agent etc., all parts team work in an entity. For the two segments, searching andfiltering, are even more important in this system, the paper amply expatiates theinformation searching Agent model and information filtering Agent model in the latterchapters. The paper brings forward the SVS searching model through comparing severalfamiliar searching model reciprocally, this model conquers the localization of the vectorspace arithmetic and puts up the advantage of the recessive semantic index model. Inaddition, the Human-based human-computer interface Agent offers intelligent querying gateways.When the websites, acquiring by ordinary searching engines, are not enough, the system will start upits own searching engine, it searches from the current websites by the breadth and depth firstlyarithmetic of the finite region, and assures the recall ratio. The information filtering Agentbreaches the oneness of the traditional filtering ways, it does the first query with the binding method of attribute, content and structure, classifies the queried information by Bayesian method, and then fetches the final results by ID3 learning method for users, it assures the precision ratio In the end, the paper tells of the user Agent, the maintenance of user file and the user database.The paper introduces the technology of intelligent Agent into the information retrieval technology. It can offer many measures of intelligent information processing neatly. It's propitious to exploiting the information service of the network information.