Application Of Data Mining In CRM Based On Customer Signature

Along with China becoming hot spots of the world economy, the developing localenterprises gradually change the mode of operation and management. They take customer as thecenter to develop product system, service system and value management system, which makesthose focus on customer segmentation，so that they can formulate a reasonable strategy to enhancetheir competitiveness. Clustering analysis and association rule model is built based on customersignature in CRM. Combined with production practice, my research shows a interpretation andapplication of the conclusions conclude by data mining，which provides a good basis for customersegmentation, personalized service and recommendation.In the practice, my paper focuses on the research of data quality control based on SQLServer DQS and the index system of customer signature according to the real production of thebakery in Hong Kong. Meanwhile, my paper also focus on designing an update plan of customersignature, realizing customer segmentation through optimized k-means cluster analysis algorithmnamely two points k-means algorithm and realizing association analys is through optimizedApriori algorithm namely Apriori-B algorithm. Bes ides, a personalized service andrecommendation has been concluded according to the conclusion of data mining. Based on theexperimental scheme in this paper, the main innovation of this paper is the attention to the effectof data quality on the final results and the update scheme of customer signature combined withhistorical environment and current environment so that it can help this enterprises to have a goodunderstanding of customers and provide reliable basis for data mining and business decision.