Design And Realization Of Adaptive Web Station System Based On Data Mining

With the rapid increase of global Web sites, the amount of information and complexities has been increased violently. How to arrange the contents and the structure of a Web site so as to enhance awareness and improve quality by employing adaptive and dynamic models for the site, has become an important research subject for Web applications.The paper addresses data mining and adaptive techniques in Web site. Firstly, transaction clustering algorithm is analyzed. The measurement for similarity of the clustering depends on discrete user browsing time instead of access time or access time. Secondly, clustering and association rule mining in Web data mining are studied. URL-UserId incidence matrix based user clustering and page clustering and frequent path mining are proposed. Furthermore, Web log mining based adaptive Web site system is designed. Adaptive site is implemented beginning with personalized recommendation and Web site optimization. Finally, Web site optimization is tested experimentally on available network environment. The structure of the site is adjusted according to experimental result.Experimental results show that the improved algorithm is simple and efficient with less needed resources. Its mining results are faithful. From the view of this point, the designed system meets the designing requirements and has some practical values.There are still shortcomings in personal recommendation and optimization of Web site for the implementation of adaptive Web site. More efforts are put on the optimization of Web site so the personal recommendation is less emphasized. The tested platform is based on a relative simple Web site of one school of a college. How to implement the system on whole campus network should be verified and studied.