Data Mining Application In Sales Management Of Retail Industry

With the development of information technology, companies have accumulated vast amounts of historical data, which become more and more difficult to manage and use for enterprises. In order to improve the data processing capabilities and get information benefit to business operations, the data warehouse can be established on the basis of traditional database for data mining, which analysis deeply from the collected data to help companies find useful patterns and rules, so that is able to make effective decisions quickly.There is a great deal of important information concealed under the mass of data relevant to sales in corporation. But at the present it can't be found out promptly and effectively because of the limitation of tool used by people. Data Mining can set up the predict model based on the sources and characteristics of sales data, helps people to pick up the useful business information from mass of data and therefore improve the market competitiveness of enterprises.This thesis illustrates the theory and technology of data warehouse and data mining, and discusses that how to use the data warehouse and data mining technology in sales management for invoicing system of golden ruyi software corporation. This thesis analyses the business activities and data relationships of the sales management, describes the design and implementation process of marketing decision support system. The thesis designs three models of data warehouse:concept model, logic model, physical model, completes the generation, usage and maintenance of data warehouse.Achieve data extraction, transformation and loading from data source through the SQL Server 2008's SSIS components. Establish a data cube using SSIS in order to meet the needs of OLAP and data mining.Based on the data mining in the application scope and characteristics of retail industry, the design framework of data mining system was proposed based on sales management module in the thesis. The framework is divided into the presentation layer, algorithms layer, and data layer. This thesis designs the B/S (Browser/Server) Decision Support System of sales management base on data warehouse in this framework, Besides the based function of sales management, it adds the subsystem of sales predict. The subsystem train and analysis a large number of data from invoicing system sales module, using BP neural network algorithm.It ultimately forecast the last three years' sales turnover, which provide decision support information for Management and decision-making of the company. In this way can effectively improve the level of sales management, facilitate the development of the company's strategy.With the development of the data warehouse and data mining technology, now more and more enterprises establish their own data warehouse. Establish a decision support system based on the data warehouse, it can improve the ability of policymakers'efficient data analysis, fully exploit and use the past scattered information. Prediction plays an important role in the future development strategy of the company.