Research On Business Analysis System Of E-commerce Based On Data Mining

Business analysis is a process of the enterprise analyzes the phenomenon which is in the course of business to make adjustment and refom timely to form the business strategy in order to control the company’s overall operation. However, coincided with the booming e-commerce activities provide a good data base for the business analysis. In the process of e-commerce, many companies accumulate vast amounts of data, these history data is10years long, but has not been good use of. In this paper, we combine with the popular data mining techniques to do the main research of the sales forecast analysis in the business analysis system in the e-commerce environment.In this paper, we combine the basic theory and methods of data mining, data warehouse and OLAP technology, departure from the development and needs of the operation analysis system, fully introduce the data mining concepts and applications, based on the status quo analysis of e-commerce and data mining and business analysis system theory, put forward the e-commerce business system design, implementation steps and so on, verify that e-commerce operation analysis system using data mining technology to improve the feasibility of prediction accuracy.In the structure of this paper, firstly we fully introduce the characteristics of e-commerce, e-commerce business system and its function; analyze the necessity of use of business e-commerce analysis system. Secondly, it describes the theory of data mining technology, commonly used algorithm and its application, and various technologies used in management analysis system, compare with data mining technology, highlighting data mining technology is superior. Thirdly, it detailed designs the structure of management analysis system, Tai Kelin e-commerce systems project as a platform, according to the characteristics of the seasonal trend in the enterprise sales data; add improved seasonal exponential smoothing method into the existing time-series data mining method, and analyze the commodity sales data to find out the laws to achieve the system.Respectively, compare the actual value with the traditional algorithm and optimization algorithm predictions, draw predictions optimization algorithm is closer to the actual value, improved the prediction accuracy, e-commerce to establish the positive significance of the operation analysis system, demonstrate the positive significance of e-commerce establishes the operation analysis system. Finally, summarize the work done in this article, put forward the lack of this article and in-depth study and improved place.