数据挖掘在电信客户关系管理中的应用

发布时间：2018-08-24 09:45

【摘要】： 客户关系管理是一套以改善企业和客户之间关系为目的的工具和软件,它将实现全面管理客户信息,分析客户行为,构建高效灵活的客户交流渠道,为客户提供完善周到的售前、售中、售后服务,提高客户的满意度和企业的赢利能力,使其在空前激烈的竞争中更好的立足和发展。客户关系管理系统主要由客户销售管理、客户市场管理、客户支持与服务管理、数据库及支撑平台等部分组成,其中在客户市场管理模块中将依托大量客户信息数据进行数据挖掘分析,以指导企业开展营销工作。本文首先介绍了数据挖掘的概念、功能、工具、流程、实现技术以及常用算法。再者阐述了客户关系管理的内涵、研究的内容以及CRM系统在电信行业的应用。随着中国电信对CDMA网络的顺利承接,为在短时间内充分掌握C网用户的消费特征,本文采集了在网用户三个月的消费数据,运用K-Means聚类算法分析了C网用户消费行为以及新业务用户的消费特征。运用Apriori关联算法分析了各业务之间的关联关系以及高价值客户使用业务的关联性。通过分析使我们明确了吴忠电信C网用户的潜在消费特征,将指导我们差异化的推广“天翼”套餐。另外,当前吴忠电信公司客户管理较为粗放,导致用户欠费居高不下,用户流失严重,为实现精细化的客户管理,差异化的做好客户欠费管理,本文利用客户资料数据、消费特征数据、欠费缴纳特征数据严格按数据挖掘的流程,经过数据预处理、探索性分析、模型建立、模型发布实现了客户信用等级评估,其中主要运用了K-Means算法、CRT算法。通过对模型的推广应用,吴忠电信运用信用等级加强客户管理,节省了大量人力资源,欠费回收得到大幅提升,客户感知进一步提高。
[Abstract]:Customer relationship management (CRM) is a set of tools and software aimed at improving the relationship between enterprises and customers. It will achieve comprehensive management of customer information, analysis of customer behavior, and construction of efficient and flexible channels of customer communication. To provide customers with perfect pre-sale, in-sale, after-sales service, to improve customer satisfaction and the profitability of enterprises, so that they can better stand and develop in the unprecedented fierce competition. Customer relationship management system is mainly composed of customer sales management, customer market management, customer support and service management, database and support platform, etc. In the customer market management module, it will rely on a large number of customer information data for data mining and analysis to guide enterprises to carry out marketing work. This paper first introduces the concept, function, tools, flow, implementation technology and common algorithms of data mining. Furthermore, the connotation of customer relationship management, the content of research and the application of CRM system in telecom industry are expounded. With the successful acceptance of CDMA network by China Telecom, in order to fully grasp the consumption characteristics of C network users in a short time, this paper collects three months' consumption data of users in the network. The K-Means clustering algorithm is used to analyze the consumer behavior of C-net users and the consumption characteristics of new business users. Apriori association algorithm is used to analyze the relationship between each business and the relationship between high value customers. Through the analysis, we make clear the potential consumption characteristics of Wu Zhong telecom C network users, which will guide us to promote the "Sky Wing" package. In addition, the current Wu Zhong telecom company customer management is relatively extensive, resulting in high user arrears, serious loss of users, in order to achieve refined customer management, differentiated management of customer arrears, this paper makes use of customer data, According to the data mining process, the consumer characteristic data and the overdue payment feature data are processed by data preprocessing, exploratory analysis, model establishment and model release to realize customer credit rating evaluation, in which the K-Means algorithm is mainly used. Through the popularization and application of the model, Wu Zhong Telecom uses credit grade to strengthen customer management, saves a lot of human resources, improves the recovery of arrears greatly, and further improves the customer perception.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2009
【分类号】：F626;TP311.13

【参考文献】