Research On Data Mining Technology In The B2B E-Commerce Merchandise Sales

With the rapid development of e-commerce and cloud applications technology, business information technology wave has swept into all walks of life. It is not a wise choice to put into a more human and material in e-commerce application software development and maintenance for small and medium enterprises. This paper is based on small and medium enterprises invoicing information management system. The system is based on SaaS (Software as a Service) and IaaS (Infrastructure as a Service) concept and design, a system of inventory management for many companies, that is E-commerce application cloud platform system. Business users buy software services, data storage and network resources on demand infrastructure services. With the increasingly fierce market competition, the most urgent needs of the enterprise are how effectively operate data for scientific, forward-looking business analysis, system analysis of business and so on.The paper analysis and data mining operational data, which is based E-commerce application cloud platform system for small and medium-sized wholesale clothing enterprises. It examines data mining and data warehouse technology, providing sales analysis and forecasting services. For e-commerce needs of enterprise customers, it identifies the structure and functions of the E-commerce merchandise sales analysis system, and present its implementation steps.This paper first introduces the concept of e-commerce, cloud technology and features, analysis the necessity of the sale-analysis-system in the E-commerce cloud system. Secondly, introduces the Data mining technology, data warehouse theory and commonly used algorithms. Then, describes in detail of the structure of sale-analysis-system. The system is based on the E-commerce application cloud platform system, and its’purpose is realizing the sale forecast and sale analysis functions. The main methods are regression analysis and clustering analysis. Besides, improving the K-means clustering algorithm depends on the initial value of K.Finally, the content of the study is applied to the electronic commerce system. Through analyzing and mining the actual sales data, in order to examine the functional and accuracy of the system. And the results is as expected, it further confirms the sale-analysis-system applying to E-commerce sites has practical significance.