The Research On Web Structure Based On Visitors

Internet monitor laboratory indicated that the types of electronic commerce website structure changes in our country at present.the website is become more multiplex in content and service.Along with the internet service developing,the work of structure improvement and the page content renew and maintenance is more complex.The website manager not only needs automatic design tool to change the structure,but also needs the user browsing information.according these information.we can analyze the layout and structure,attract more visitors.In this research,we explain the concept,research works,key techologies of Web log mining,difficulty and related research at home and abroad,and the use data mining technology to analyze the Web usage information of service website in Shanxi so as to find out the usage pattern and preference of enterprises and individuals as the better decision-making aid for website manager.the paper achieves the following tasks:At first,in the current foundation,stydying the preprocessing of raw Web log,analyzing the difficulties and describing the process;second,on the basis of the introduction of the cluster analyze, the usage of it.cluster the web log dates;third,using these date,we can discovery of sites'visitors browsing behavior,the improvement of sites'structure and the linkage structure and the decision support on client relationship management of e-commerce.;finally,depending on the adjust advise,we chang the web structure,prove the research rationality.