Research On Website Usability Analysis And Visualization Technology

With the Internet developing rapidly, more and more information resource are being used. The information for people communicating inevitably got large and electronic. It is too difficulty to find information people interested in from huge and unorganized information. Web mining technology provide a new way to solve this problem. Use Web mining technology to analyze huge Web usage and do research on visualization technology of analytical results in order to let users find information they need easily and conveniently, accordingly improve the Website Usability.Firstly, the paper introduced research status at home and aboard, especially emphasis on analyzing and discussing date preprocessing procedure. Bring forward a Web usage mining preprocessing arithmetic which carry out user identification, session identification and path complement at one time, avoided former division of algorithmic execution and read huge Web log dates repeatedly, so it improved efficiency. Secondly, analyze clustering arithmetic and ant algorithm theory, then bring forward a Web users' visiting path clustering arithmetic found on ant algorithm which classify the group by the most frequent visiting page, consider visiting time and how important the user is , integrate with user visiting sequence, whether have same father or child page to account the path similitude. For the whole clustering process, adjust the clustering group number dynamically to access fine clustering results. After clustering, find association rules based on that. Finally, analyze Web visualization technology and visualized Web usage mining results so that the Website manager can find Website usability problem directly.On the experimental results, preprocessing arithmetic the paper forwarded is effective, improve the date quality for mining to deal with, accordingly advance the efficiency. Web users' visiting path clustering arithmetic found on ant algorithm can reflect users' visitingpattern more accvirately. The personal recommendation according to transfer probability of ant algorithm achieve higher including rate and accurate rate than Apriori arithmetic and can reflect users' visiting trend correctly. Meanwhile, Web site manager can refer the results to adjust the site linkage structure and provide individual service, accordingly improve the Website Usability.