Research And Implementation Of Data Analysis Method For Traditional Retail Trade

With the development of science and technology and the popularization of information technology,retail enterprises have accumulated a large amount of user information and historical consumption data which hides a lot of valuable information.In today's diversified market environment,the traditional retail trade must make full use of these data in order to maintain vitality in the fierce market competition.How to use the relevant technology to dig out valuable content from the data to guide the business activities of enterprises and improve the decision-making ability has become a critical issue that traditional retail trade must face.At present,there are still many problems in the use of consumer data in many retail enterprises.First,the original amount of data is too large and complex to locate high quality information which causes a lot of meaningless data interference.Second,a lot of work is still in the basic data browsing,query,statistics and other low level operations.Operators are unable to dig deeper and get more valuable information.Third,due to the lack of powerful analys is tools and data mining experience,the effic iency of data mining is low in traditional retail trade.All of the above limits the use of data for retail enterprises.In this paper,the above issues will be studied in depth by exploring these data,combined with the characteristics of the retail trade to find out the hidden characteristics of consumer and shopping behavior.The main research objects include consumer's personal information and commodity sales data.The main contents are as follows.First,studying the related data mining technology and methods,includ ing cluster analys is and association rule mining.Second,pretreating the original data,inc luding extraction,cleaning,replacement and merger.Third,analyzing the characteristics of statistical data from the two dimensions of user information and commod ity sales.Fourth,proceeding cluster analys is on customers based on several important attributes of users.Fifth,mining association rules according to the user clustering results and summarizing the experimental results.Based on the above work,this paper deals with a complete set of processes for the retail data and gets some valuable results.This information will help retailers to improve business decisions and corporate revenue,and also provide a viable research direction for the future work.