Research And Application Of Web2.0 Personalized Service Based On Data Mining

A new Internet revolution has been coming to us due to the outcoming of Web2.0 in which Long Tail theory has replaced the status of 80/20 rule as the basic principle of the advance of Internet Business. As to Web2.0, it has a lot of divergence with Web1.0. For instance, Web2.0 focuses on the concept of "everything based on user", which therefore determined the primary objective that Web2.0 takes pain to pursue would definitely lie in the issue: by what means could Web2.0 serves the purpose of seizing users'interests and thus to offer personalized service. Accordingly, it is urgently needed to research on the Web 2.0 personalized service and its applications.The first part of this paper introduces the concept and background of Web2.0, pointing out its difference compared to Web1.0. Later, we introduce Long Tail theory——the basic principle of Web2.0, and also the challenges that it puts on Web2.0 to render personalized service to user with regard to ameliorate user experience.To meet such demand, our paper introduces clustering method of data mining. By pointing out the limitation if diverse modeling method, we come up with our user clustering modeling. Essentially, it is an abstract modeling approach. As for this paper, we adopt K-means algorithm as the key part of user clustering. Theoretically and experimentally, we validate the advantages of user clustering modeling.Besides, we further our research on applications of user clustering modeling. By means of embedding related modules of user clustering modeling method, a personalized program recommendation system is impelmented. In contrast with the result of applying statistical method, the superiority of user clustering approach is evidently exposed and further proved by an instance exposition.At the last part of this paper, we conclude our research and come up with future work.