Research On Mobile Agent Based Personalized Information Retrieval

The thesis recommends the concepts, relevant technology of Mobile Agent and information retrieval, and trends both at home and abroad at first, describes the MABIRS information retrieval system that use the mobile agent and feedback technique.the paper expatiate the retrieval framework, explain the basic function of index database, user interest tree, user interest datebase, user interest module, classification, Mobile agent manager and every kinds of mobile agent. And introduce the system' s realize theory, system framework and the relation of user interest datebase, index database, user interest module and user interest tree, and combine with the modules function flow chart state the system' s design process.In the paper, we design a three layers security module based on inspection module, three net classification principle and correlation adjust arithmetic based on user' s interest. We adjust correlation in the user' s interest module use all kinds of parameter correlative with user' s feedback, and appraise correlation between net and sort. Via correlative experiment, prove the method is validity, and establish the base for set up effective personalized information retrieval system.