The Research Of Personalized User's Profile Based On Web Mining

With the development of the Internet and information technology, the problem of"Information Explosion" has arisen, that is, "Rich Data and Poor Information". How tomanage the tremendous mount of information on WWW to meet the growing needs ofpersonalized information is a new subject for our research. Personalization has been the focusof research. Personalization, that is to give different service-strategy and differentservice-content to different user. Knowledge of user interests and describing them by userprofiles are the importance. Therefore, the effectiveness of personalize information serviceprovided by the system is determined by the fact whether or not the user profiles reflect userinterests exactly.After the studying of the key technologies――web mining technology and modelinguser profile, the paper suggests the model of mining user interests. The model is based on userviewed content and combining with analysis of user's behavior. Through analyzing documentexpressive model, feature extraction and feature weigh value, the web page is been expressedby Vector Space Model.Then, in this paper we do hard in two aspects: clustering based on content, creating theuser interest model. After the probing into the cluster algorithm existingand the appliedpractice, we proposed CLOPE algorithm. After get the cluster, we use tree-model to expressthe user interest, it is as ( I₁,InterestDegree(I₁)),( I₂ ,InterestDegree(I₂ )),…,( I_k ,InterestDegree(I_k ))).Finally, the paper experiment on the advanced method discussed above. According to theexperimentation and analyses, prove that tree format interest model is reliable, and can beapplied in personalization system.The future work of this paper is that developing the validity of user interest model, andapplying it into the recommendation.