基于数据挖掘技术的超市顾客群研究  

Study on Customers′ Group of Supermarket Based on Data Mining

作　　者：肖生苓[1] 牟娌娜[1] 王维[1] 高晓红[1] 

机构地区：[1]东北林业大学工程技术学院,黑龙江哈尔滨150040

出　　处：《资源开发与市场》2011年第8期683-685,712,F0002,共5页Resource Development & Market

摘　　要：利用SPSS Clementine 10.1数据挖掘工具,遵循CRISP-DM模型的商业目标、数据理解、建立模型的流程对超市顾客进行分析研究。提出衡量超市客户忠诚度的忠诚度系数指标,建立忠诚度—盈利性顾客细分模型,运用k-means算法对超市顾客进行聚类分析,帮助超市企业准确识别不同类型的顾客群,尤其是忠诚的高盈利顾客。再利用所建立的序列分析模型分析顾客类别变化路径,预测顾客价值变化趋势,及早发现潜在价值顾客,使其尽早成为企业忠诚的高价值顾客,实现超市企业利润的有效提升,最终在日益激烈的商业竞争中立于不败之地。Using data mining tool of SPSS Clementine 10.1,this study analyzed the customers′ group of supermarket based on the procedures of CRISP-DM model-business objectives,data understanding,model buliding,and model evaluation.A loyalty coefficient index was pointed out which measured the loyalty of supermarket customers.A customers subdivision model of loyalty-profitability was then developed.Clustered analysis was carried out towards customors of supermarket using k-means algorithm to aid the supermarket company,and to identify different types of client accurately,especially the customers who could bring high profits to the supermarkets with loyalty.The variation of customers was analyzed based on the sequence analysis model and the trend of variation of customer value was then forecasted to help the manager of the supermarket identify the potential valuable customers.The potential valuable custormers could be developed into loyal customers with high value to the supermarket in order to achieve the high profitability and high competitiveness.

关 键 词：超市 顾客分析 数据挖掘 聚类分析 序列分析 

分 类 号：F272.5[经济管理—企业管理]