Research And Implementation Of Customized E-commerce

As Internet becomes more and more widespread, E-commerce is affecting the development of enterprise and life of the people significantly. Internet is a world of information, and in this information's world, users usually receive a great deal of useless information reluctantly while search useful information. This not only wastes users' time and energy, also makes many users lose patience to visit Internet. This is no other than the shortage of traditional E-commerce. According to these drawbacks,traditional E-Commerce are prevented from developing further.The customized E-Commerce offers a very promising solution for the problems in the current E-Commerce. Usually when visiting a web site, users want to browse some information they need, and the businessman also want to offer content that users need.In this new pattern of E-Commerce, when we establish a web site, we must analyze users' characters and customize corresponding web pages, and show different information or advertisement to different users. This paper works over users classification in the customized E-Commerce mode. First, I analyze behave information of browse, sample the data of every users, and find classification of the samples, so find correspond class of some user. Then, according profile information which users register when first log in, use a structure of tree, and propose a new association rule algorithm based users' profile information. This algorithm deal with quantitive attributes to build rule sets, and cluster the rules. Moreover I implement a B to C web site based on this algorithm.