Data Mining Application In Tobacco CRM

In recent years, along with the society to the tobacco control attention degree increase,tobacco control efforts continue to increase, the pace of development of low tar cigarettesgradually accelerated, low tar cigarette sales will be the main trend of tobacco sales in thefuture. Facing the hitherto unknown competition pressure, retail terminal as linked tobaccosystem and consumers a bridge, can be said to be the key to the competitiveness of tobacco.Therefore, how to excavate the potential high value retail customers, promote thedevelopment of cigarette retail customer business development direction to move closer tothe tobacco industry is the priority among priorities in the tobacco network constructionwork.The service from the tobacco industry customer strategy, analysis of type CRM is thefuture trend of development, through analyzing the operation of CRM data extraction,analysis and prediction, the large amount of data into a reliable and practical information,guiding the tobacco enterprises in cigarette sales and customer service aspects of theallocation of resources, and ultimately improve customer the purpose of the relationship.Tobacco commercial enterprise has accumulated a lot of customer order data in the course ofbusiness, the customer order data was analyzed by using the data mining technology,customer segmentation, thus the implementation of personalized services of high valuecustomer training with large sales potential, in low tar cigarette, make customer developmentand mutual promotion of enterprise development, customer profit and enterprise benefitmaximization.This paper is based on the Microsoft SQL Server Analysis Service (SSAS) and datamining, mainly uses the B/S architecture based on Web, including data source, datawarehouse, OLAP, application server and client. According to the actual demand analysis oftobacco marketing, to build a data warehouse retail customers based on the theme of cigarettesales in SQL Server2005, and from the source database, convert and import data into the datawarehouse. Then based on data warehouse multidimensional data analysis to establish thecorresponding SSAS on the theme set, using DMX language to achieve the drill, slice, dice various demand analysis and data, and developed with the show in front of Web data basedon Microsoft Reporting services.As the data source into an area of Guangzhou tobacco’s historical sales data of theresearch to the SQL Server database, and use SSAS as a data mining platform, construct datamining model, and data mining using the decision tree classification technology, realize thecigarette sales trend of decision analysis. And, through analysis of customer sales data miningof cigarette retail customer sales potential, and sorted to the potential customer value,cultivation of high value to find potential customers.