Based On Intelligent Agent Web Personalized Information Retrieval System

Traditional search engine can not meet people' s information needs. Facing the ocean of information on Internet, people always get too much information to select and to digest, so it' s hard for them to find the necessary information and knowledge. Web mining is a new technology and it can discover resources and knowledge quickly and efficiently, therefore, it can compensate the shortcomings of search engine.The aim of the thesis is to help people who search and mine domain information on Internet. In view of the question that existing search engines cannot express requirements of users exactly and output too many results, based on the analysis of the history of Internet and Internet Information Retrieval, then a system of Web personalized information retrieval based on intelligent Agent is presented by intelligent Agent technology. By the description of the Web information retrieval workflow under this system, the paper briefly explains the overall thought. The system also discusses initially abou designing system and building model. Around the building and software development of the model, the main work and innovation of this thesis are listed as the following:Firstly, the presented model has a multi-agent architecture and orients the practical user. In the model system, user and document representa _tions are depicted. Documents are the objects of information process and the foundation of document model is the base of information process such as feature extraction, document filtration and so on. User model embodies user interests and intention. It is used to express and mining user interests.Secondly, in the information retrieval agents, a key problem is how to extract text features and how to deal with the high number of the feartures. retrieving-monitoring agent uses the feature vector to search information and document on Internet. Analysis-filtrating agent extracts...