Data Mining Research In Tobacco E-commerce

With the flourishing development of the Internet and computer network technology, network and globalization have become the irresistible trend of the world. The application of e-commerce business, improve the tobacco trading platform, strengthen the exchange and communication with retail dealer , have become the powerful means of the tobacco enterprises to enhance the core competitiveness and promote the enterprises rapid development.Data mining is a newly emerged cross discipline.It extracts information and knowledge which it is not implied but potentially useful from mass,uncompletely,noisy and fuzzy of practical application data .It take the application of data from the lower levels of simple query operation, upgrade to mining the advanced application of useful knowledge and information resource from the mass data resource. It can provide information support for enterprises' comprehensive decision-making .This paper designs to construct a tobacco e-commerce platform, and applicant the idea and method of data mining into e-commerce. According to the demand characteristics , purchasing behaviors, buying habits and other aspects of retail dealer, through the study of mass, complex and dynamic dealing data, using the association rule Apriori algorithm of data mining to achieve related sales and brand recommendation and other functions of cigarette products, and guide the shelf display of cigarette retail dealer. On the other hand, introduce the k-means algorithm of cluster analysis by e-commerce transaction data, and based on the segmentation of retail dealer values and subdivision of retail dealer behaviours, to establish retail dealer segmentation models and constitute the retail dealers of similar characteristics into a same group.Then comprehensively evaluate and analyse retail dealer, develop targeted marketing programs, make resources rationally allocated, realize customers precision sales, and promote enterprise's overall operational level, establish tobacco e-commerce strategy regarding customer as the center .