Research On Web Data Mining In E-Commerce

With the popularization of Internet and the development of E-commerce, E-business Web sites are faced with more and more fierce competition. The method that focuses on SCD(Web site centered design) will be replaced with that focuses on UCD(user centered design). A key problem with E-business development is: how to make the best of the plentiful E-commerce information and mine the user interest pattern to understand user behavior, improve the'usability of web site and provide personalized service for customers.Web mining is the extraction of interesting and potentially useful patterns and implicit information from artifacts or activity related to the World-Wide Web. Web data mining has played important roles in many fields. E-commerce provides Web mining a new task and abundant data resources.First, we find that there are plentiful data in E-commerce haven' t been used well. On the basis of this, the thesis proposed a new method to mine for user interest patterns and the favorable pages, which fully considers user browsing time and switching among Web pages. Second, Aimed at the main challenges of recommendation algorithm in E-commerce, this thesis proposed a product catalog-based multi-level association recommendation algorithm. Combining user browsing information and product catalogs, the algorithm can make recommendation based on frequent item sets. It greatly enhances online performance and quality. In large E-commerce systems, the real-time requirement of collaborative filtering recommendation system is hard to be satisfied. To address this issue, we proposed a collaborative filtering algorithm based on clustering. It can improve the performance of commendation. At last, a framework of recommendation system based on Web mining is proposed.