Personalized Research Combining With Analysis Of Web User's Behavior Based On The User's Browsing Content

The rapid improvement of Internet and World Wide Web makes the design andmaintenance of Web sites more and more important. How to manage the tremendousmounts of information on WWW to meet the growing needs of personalizeinformation is a new subject for our research. Personalization has been the focus ofresearch. Personalization, that is, give different service-strategy and differentservice-content to different user. Knowledge of user interests and how to describethem by user profiles are the importance.After analyzing the technology of data mining, further, the author studied the keytechnologies ――Web mining technology and modeling user profile, and the authorsuggests the model of mining user interests. The model is based on user viewedcontent and combining with analysis of user's behavior. Through analyzing documentexpressive model,feature extraction and feature weigh value, the Web page is beenexpressed by Vector Space Model.In the paper, the author did hard in two aspects: clustering based on content, creatingthe user interest model. By the manner of similarity computing with the different texts,author did the analysis of algorithm. After the probing into the cluster algorithmexisting and the applied practice, the author proposed a new cluster algorithm:combining agglomerative algorithm with K-means algorithm. In the process of cluster,use the agglomerative algorithm to get the cluster-means and k firstly, and then useK-means algorithm to do the second cluster. After get the cluster, the author use twolevel tree-model to express the user's interest. For the sake of using and updating ofuser interest model, every interest style of user is express by VSM as the Web page.So, the compare of Web page with user interest style can be valued by similarityfunction. Finally, for making the theory objectively and specifically, the authorexperiment on the advanced method discussed above.According to the experimentation and analyses, prove that the new cluster algorithmand tree format interest model are reliable, and can be applied in personalizationsystem. Lastly, the deep work of this paper is that developing the validity of userinterest model, and applying it into the recommendation.