携号转网背景下的客户满意度提升研究
首发时间:2020-04-10
摘要:在携号转网背景下,运营商企业将面临客户流失的风险。从全量客户中筛选出有不满意倾向的高危群体,并提前对这些客户进行关怀和维系,具有非常重要的意义。论文基于客户历史行为数据,分别构建了基于卡方决策树、C5.0决策树和Xgboost森林的客户满意度评估模型。测试结果显示,比较准确性而言,Xgboost模型>C5.0模型>卡方模型;比较执行效率而言,C5.0模型≈卡方模型>>Xgboost模型;综合而言,C5.0模型最优。论文给出了基于C5.0模型的不满意客户的画像特征,并基于该画像特征输出了分层分级的客户关怀策略。
For information in English, please click here
Research on the improvement of customer satisfaction under Mobile Number Portability Background
Abstract:Under the background of Mobil Number Portability, operators will face the risk of customer loss. It is of great significance to select the high-risk groups with dissatisfaction tendency from the whole number of customers, and to care and maintain these customers in advance. Based on the historical behavior data of customers, this paper constructs the evaluation models of customer satisfaction based on chi-square decision tree, C5.0 decision tree and Xgboost forest. The test results show that the Xgboost model >C5.0 model >CART model is more accurate. In terms of execution efficiency, C5.0 model ≈ CART model >>Xgboost model; Overall, the C5.0 model is optimal. This paper presents the portrait characteristics of dissatisfied customers based on C5.0 model, and outputs a hierarchical customer care strategy based on the portrait characteristics.
Keywords: Mobile Number Portability Customer satisfaction Xgboost The decision tree
基金:
引用
No.****
同行评议
勘误表
携号转网背景下的客户满意度提升研究
评论
全部评论0/1000