电信客户细分中基于聚类算法的数据挖掘技术研究  

Research on data mining technology based on clustering algorithm in telecom customer classification

作　　者：梁霄波[1] LIANG Xiaobo(Guilin Normal College,Guilin 541001,China)

机构地区：[1]桂林师范高等专科学校,广西桂林541001

出　　处：《现代电子技术》2016年第15期95-98,共4页Modern Electronics Technique

基　　金：广西职业教育教学改革项目一般项目(桂教职成[2015]22号)

摘　　要：结合市场细分和数据挖掘技术等相关理论,针对电信行业客户细分领域存在的问题,以某市电信公司提供的二十多万客户数据作为研究对象,基于CRISP-DM数据挖掘模型,通过数据挖掘工具中提供的K-means聚类算法完成对电信市场客户准确、有效的聚类细分,为管理和营销人员初步提供了制定管理或营销计划的依据。在完成客户细分模型的基础上,设计和实现了电信客户细分系统,并编码实现该系统功能的主要模块,将客户细分的整个流程通过自动化的过程描述出来。According to the market segmentation,data mining technology and correlation theories,and aiming at the prob?lem existing in customer classification field of telecom industry,more than 200 000 customers′ data provided by the telecomcompany of a certain city is taken as the research object. On the basis of CRISP?DM data mining model,the K?means clusteringalgorithm in data mining tool is used for achievement of accurate and effective clustering classification of the telecom customers,which provides a basis of making a preliminary management plan or marketing plan for the management and marketing staff. Onthe basis of the customer classification model,the telecom customer classification system was designed and implemented,andthe main module of the system function was realized with encoding. The whole process of customer classification is described bymeans of automation process.

关 键 词：数据挖掘 客户细分 SPSSModeler K.means算法 

分 类 号：TN911.34[电子电信—通信与信息系统] TM417[电子电信—信息与通信工程]