Potential Customers Mining In E-commerce

In recent years, e-commerce (Electronic Commerce) showed a rapid growthmomentum and gradually moved toward robust and mature direction. Consumers areincreasingly dependent on the network for consumer shopping. Thus, e-commercecompanies mine potential customers that can increase enterprises’ profit andefficiency. That is, who can grasp the needs of the customers, improve thecompany’s image and do well in the relationship with the customer, as well aseffective mine customer resource management, who will be able to increase marketcompetitiveness and gradually increase its market share in an invincible position.How to mine potential customers has been the focus of E-commerce companies’competition, which is one of the key factors for their success.The E-commerce market has been increasingly competitive and getting morerelated information of consumers is of great significance for the business companies.From the back-end database of the E-commerce, we can rich customers’ transactioninformation and related data. The data is very large amount of information, despite itcontains a lot of useful information about customers. However, due to the factors ofredundant, unclassified and disorganized data, we are not able to get full use of thedata. As a result, how to extract useful information from these data to mine potentialcustomers has been the key problem. In order to solve this problem, Web data miningtechnology come into being. We mainly discuss how to implement the function ofmining potential customers by study Web data mining in this article.From the perspective of information acquisition, data preprocessing, user clusteringand so on, this paper discuss the application of data mining algorithm in miningpotential users. First of all, from the perspective of the theory, we try to understandthe basic knowledge of what is potential users, data mining, classification, user accesspath, etc. Secondly, in-depth study the pretreatment process of Web log mining,including data cleaning, user identification, session identification and transactionidentification. Then, further study Kohonen neural network algorithm, by using c++programming language to implement it, and using experiments to illustrate that theimplementation is correct. Finally, give the prototype of mining system that isdesigned for potential group of users based on Kohonen neural network algorithm,and give the preliminary implementation of user mining.