The Research And Design Of Data Mining System Based On E-Commerce

In recent years, as people rapidly improve their ability to utilize the information technology production and collect data, thousands of databases are used in business manager, government's official bussiness, scientific research and project development etc. But at the same time, many data meet a serious problem: proliferative data have many important information behind, but people how to make beter use of these data avoiding phenomena of "explosive data and poor knowledge"?The data mining technology is the new hot spot in the computer world. Through the effective excavating of a large amount of data in history, can find the potential knowledge in thes data which may be very helpful to the decision support, but these modes can not rely on the simple inquire, or can not get within acceptable time. These modes become knowledge under the discernment of professional person. It is complicated that the tasks of data mining usually include classifying, predicting, association rules discovery and clustering, etc.With the rapid development of Internet, e-commerce system is used more and more widely in nowadays life. Its function has extended from online transactions to all kinds of business on the network. The building technology of the e-commerce system should not only support the development and reuse of system, but also provide the ability of intelligent analysis based on data mining and other technologies, which can help the enterprise to improve their customers, relationship management and marketing activities. There are some questions should be researched on. How to improve the reusability of software? How to build the development platform of e-commerce system based on reusable components and design paterns? How to apply data mining into e-commerce system? Researches on these questions are proved to be valuable. This paper is mainly about these questionsThis paper discuss two parts as follow:1. At first introduce basic conception of Data mining and relationship with data warehouse and OLAP. Then, we combine e-commerce and set up the scheme of customer's behavior history based on Microsoft SQL Server 2000 and the OLAP technology, offer the clean data for data mining based on the algorithm of decision tree.2. In this paper, we introduce the prototype of data mining system based on the e-commerce. Then, we get the decision tree by the ID3 algorithm theory through the OLAP data set. And decision tree was directly realized in client programming.This provides some reference for research of other data mining methods.At last we give the main contributions of this paper and our future work as a summarization.