Btob Website User Access Pattern Mining

With the explosive growth of information sources on the World Wide Web, it has become increasingly necessary for E-commerce companies, who have transfer the Web into the primary tool for its development, to utilize data mining tools to find, extract and analyze user access patterns in their E-commerce websites. In this paper, we combine the data mining technology and E-commerce websites effectively and make it a useful method to analyze and evaluate E-commerce websites.This paper is organized as follows. First, we summarize the data mining technology and introduce E-commerce system. After having known the above two, we introduce the data preparation. Then we analyze mining algorithm detailed in two aspects, one is the sequential pattern, and the other is the association rule. In sequential pattern, we describe MFR and the algorithm on how to find frequent access paths. In mining association rules, we introduce famous Apriori algorithm and propose the optimized DHP algorithm with hashing. Afterwards, a quick method and its optimized eclat algorithm without creating candidate are presented. Then, we do experiments with the algorithms in a website. Finally, we conclude our work with a short summary and point out some research directions on this work in the future.