The Research On B2C Website Usage Pattern Using Data Mining Method

By applying the method of data mining in e-commerce and web database, the system could afford intelligent and personalized service for everyone. Taking the advantage of the data mining, we could optimize the e-commerce website structure, have a good system performance, discover the business intelligence and recommend personalized page and product to customers. With the help of web data mining, we can analysis the visitors'behavior and interests, thus extract the special features of each visitor group. With this important knowledge, we could have better understand on these visitors, and take related promotion strategy on them.The dissertation mainly focuses on the application of using data mining method in e-commerce website. The main work includes:Designing a data minding oriented e-commerce website structure, based on the analysis on e-commerce website requirement. The structure itself is structure optimized and could advance the mining efficiency.Suggesting a method to make pre-process on site visiting log file. It could help to find the visitor's transactions to form the database for usage pattern mining. By combing the MFP method and an improved mining algorithm, the efficiency of usage pattern mining could be higher.The method of content mining and discovering related product in e-commerce website is also introduced. By using the collaborative filtering method, and customer classify, the system could make personalized web page recommendation to different users.Based on the research above, the author made the programming and test for the main part and do the first step test on a certain B2C website. The dissertation gives an initial guideline for building a personalized and intelligent e-commerce website..