欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
电信行业是典型的数据密集型行业，拥有大量的甚至是海量的客户数据资源。对电信行业客户消费数据进行深入挖掘可以为企业的资源优化配置和客户关系管理提供理论支持和技术保障。以电信行业的客户消费数据为基本研究对象，在衍生特征构造、样本调整以及特征选择等数据预处理的基础上，本文采用可处理混合数据且具有近似线性时间复杂度的一趟聚类算法建立电信行业的客户细分模型。经实证研究表明，该模型可以将电信的客户有效划分成四个具有不同忠诚程度和消费能力的客户群体，同时从各客户群的消费行为中还可以有效地分析出他们的消费偏向和流失倾向。说明提出的方法是一种有效的客户细分方法。
英文摘要：
Telecom is a typical data-intensive industry, with large or even more amounts of customer data resource. It can provide reliable theoretical and technical support for the resource optimal allocation and customer relationship management by the depth mining of customer＇s consumption data. On the basis of data preprocessing, such as feature building, sample adjusting and feature selection, one-pass clustering which has nearly linear time complexity and can handle mixed data, is used to build customer segmentation model for telecom customer data. Empirical results demonstrate that the model can divide the customers into four different groups with different loyalty and consumption ability, From the consume behavior analyzed of customer groups, it indicates that the established model is an effective customer segmentation method and can effectively identify the consumption preference and propensity tendency of different customer groups.
同期刊论文项目
同项目期刊论文
期刊信息
位置：