摘要
大客户是各移动运营商利润的主要来源，也是竞争的焦点。随着市场竞争的日益激烈，如何降低大客户离网率，是摆在各运营商面前的战略性任务。采用数据挖掘技术，遵循数据挖掘标准流程CRISP-DM。从商业理解、数据理解、数据准备、建立模型、模型评估和结果部署等6个阶段，详细介绍了移动通信企业中大客户离网预测模型的建立过程和方法。同时对预测结果从技术和业务上进行深入分析，以辅助运营商及时采取措施进行挽留。
High-Level customers are the main source of profit of every mobile communication provider, and are the focus of the competition between these providers. With the increasing of the competitive market, how to control the customer leaving is a strategic task for the providers. According to the standard data mining process, CRISP-DM, the procedure of a high-level customer churn prediction model is discussed thoroughly from the six phases of business understanding, data understanding, data preparation, modeling, evaluation and deployment. And more the predictive results from the technology and business view are deeply analyzed in order to help reduce the high-level customer leaving.
出处
《江苏通信技术》
2004年第3期1-4,共4页
Jiangsu Communication Technology
