基于数据挖掘的大学生客户识别模型的研究

【作者】 陈思；

【导师】 赵晓侠；

【作者基本信息】 昆明理工大学 ， 仪器仪表工程（专业学位）， 2015， 硕士

【摘要】 随着移动互联网的高速发展,电信业务的不断推进,用户数据的不断累积,大数据时代的来临,整个通信行业的竞争越来越来激烈。客户数据资源积累成为左右行业营销策略导向的重要资源。电信行业作为用户基础通信数据的核心载体,随着4G(第四代移动通信技术)网络的不断普及,用户作为利润增长的核心要素成为各大通信运营商的争夺焦点。作为客户来源的主力军,年轻用户通信活跃度不断提升,对通信运营商市场份额起到至关重要的作用。而大学生客户作为年轻用户的代表,对这部分人群进行行为特征分析具有十分重要的意义。本文以中国移动K市的校园区域基站下的客户基本信息为依据,按照CRISP-DM标准过程模型的商业理解、数据理解、数据准备、建立模型、模型评估、模型部署共六个阶段内容进行数据挖掘工作,选用数据挖掘工具SPSS Clementine12.0.3,利用决策树C5.0算法、神经网络算法、二项logistic回归算法对大学生客户识别模型进行建立并评估,选择最优模型,并验证了模型的唯一性,进而对模型进行部署。在此模型的基础上,业务人员根据大学生客户的通信情况制定有效的营销策略,更好地为大学生客户服务。更多还原

【Abstract】 With the high-speed development of mobile Internet, telecommunication business steadily, user data continue to accumulate, the big data era, the pattern of the whole communication industry competition is more and more fierce.Customer data has become an important resource for the accumulation of resources around the industry-oriented marketing strategy.The telecommunications industry as the core user base carrier communication data, as 4G (fourth generation mobile communication technology) growing popularity, the user as a core element of profit growth has become a major focus of contention communications carriers.As the main force in the customer base, the young user communication activity rising, the market share of the telecommunications operators play a crucial role. And college students on behalf of clients as young users, this part of the crowd behavioral analysis has important significance.In this paper, the basic information about the campus area customer base K and municipalities of China Mobile as the basis, in accordance with standard commercial understanding CRISP-DM process model, data understanding, data preparation, model building, model evaluation, model deployment of a total of six stages of the content data excavation work, the choice of data mining tools SPSS Clementine12.0.3, C5.0 decision tree algorithm, neural network algorithm, two logistic regression algorithm model for college students to establish customer identification and assessment, to choose the best model, and validate the uniqueness of the model and then the model deployment.On the basis of this model, the business personnel according to the communication situation of the university students’customer to develop effective marketing strategies, to better serve customers for college students.更多还原