On Ec-Oriented Web Data Mining

Data Mining is a kind of fairly new information technology that has been developed with the technology of database and Artificial Intelligence. Data Mining integrates knowledge from many subjects such as database, AI and statistics, tring to extract the unknown, effective and useful knowledge from data. Especially, with the popularization of Internet, the data on the Web is increasing dramatically at the speed of 20% per month, thus Web has been the most important media for storing, distributing and aquiring information. How to find the useful information behind the huge data and improve its usability, but not be drowned in it? Web Data Mining is a powerful solution.Web mining is the traditional data mining technology used in Web, which can extract user' s browse pattern from data on Web. Web Mining is devided into three groups: Web Content Mining, Web Structure Mining and Web Usage Mining. Of which, Web Usage Mining can get the interesting pattern from the data user browsed, and apprehend the user' s browse interest behavior, in order to improve the Website' s structure and provide individual service for the user.This paper introduces Data Mining and Web Mining technology, including their definition, classification, working flow, applications as well as their developing trends. Especially we analyze E-Commerce, including its personalized recommendation system, its data sources, its related technologies and its application fields. At last we design a practical recommendation system based on Web Usage Mining, which can extract user' s browse pattern from Web log, providing personalized service to different user. In this system, we discuss all the moduals and give its whole working flow: from data collecting to data preprocessing to off-line Web Usage Mining to on-line recommendation and so on. We study three technologies in Web Usage Mining in detail: Data Preprocessing technology, Pattern Discovery technology and Pattern Analyzing technology.