We also see the need to rationalize different channels of service provision - counter , fax , email , post , phone , online , etc . we need to segment our customers and understand what they want and to deliver services in a manner that best suits them in a cost - effective way , through a proper channel management strategy 我们也需适当调整提供服务的不同途径柜台传真电子邮件邮递电话互联网等。我们需把客户分类，了解他们的不同需要，然后透过适当的服务途径管理策略，以符合成本效益和最切合客?需要的方式提供服务。
The paper firstly introduces crm , bp ann ( artificial neural network ) and the traditional methods of customer classification , then studies the customer - classification method according to customers " value and improves on the existing customer value model . subsequently , applying the bp ann to customer classification , the paper puts forward the customer classification method based on bp ann . in the end , the paper introduces a crm system developed by the author during the past year 本文在分析客户关系管理、 bp网络、传统客户分类方法的基础上，探讨了以客户价值为依据的客户分类方法，对现有客户价值评估模型加以改进，提出了新的客户价值评估模型；考虑到客户价值难以直接计算，将bp网络应用于客户分类，提出了基于bp网络的客户分类方法；并开发了连锁超市客户关系管理系统。
Credit risk management is subject concerning a large amount of fields , and with limitation of my capability and the length of this paper , i specialize in the control before enterprises ’ account sale which includes information investigations of the customers and their classification and how to ascertain the amount of account sale 企业信用风险管理是一个宏大的研究课题,由于能力和篇幅的限制,本文着重探讨了企业赊销前信用风险的控制,包括企业客户信息调查和客户分类以及赊销额度的确定两个方面。
We also see the need to rationalize different channels of service provision - counter , fax , email , post , phone , online , etc . we need to segment our customers and understand what they want and to deliver services in a manner that best suits them in a cost - effective way , through a proper channel management strategy 我们也需适当调整提供服务的不同途径柜台、传真、电子邮件、邮递、电话、互联网等。我们需把客户分类，了解他们的不同需要，然后透过适当的服务途径管理策略，以符合成本效益和最切合客?需要的方式提供服务。
Secondly , according to the customer ' s habits and consume concept , we propose that customer ' s consume concept should be changed in order to enlarge the profit of management . thirdly , the customer satisfaction model is discussed . as for retaining and upgrading of credit card customers , the realization of customer lifetime value is discussed and it is proved not to be a good standard for customer segment in reality 对于信用卡客户的维系和提升，本文在分析客户终生价值作为客户分类标准，应用于实践的不可实现性的基础上，提出了通过客户的钱包份额、信用卡用户的还款方式、客户实时信用程度三个方面的细分的客户分类，并根据每一类客户的特征做出了相应的差别化管理策略；同时，提出了信用卡用户的客户终生价值的组成部分，作为信用卡客户价值提升的参照。
At last , we put much efforts on introducing our two data mining tasks we have completed on the consumers " behavior data ? ? mining association rules based on the apriori algorithm and prediction the class of the registered consumers by means of build decision tree based on the sliq algorithm 最后，我们着重介绍了对数据集市中的数据所进行的两项数据挖掘任务? ?第一个是在apriori算法基础上实现业务关联规则的挖掘，另一个是在sliq算法基础上实现创建注册用户信息决策树以进行客户分类预测。
( 4 ) build a " pyramid " customer model and propose a new formula of customer satisfaction with making classification of history customers and refinement of customer according to the customer ' s characteristics . with different processing for different customers , customer satisfaction is increased perfectly and redevelopment business chance is obtained successfully ( 4 )构建了“金字塔式”的客户模型，对掌握的历史客户分类，并按照客户的关键特征对客户类别的细分进行研究，并提出了一种新的满意度计算公式，针对不同的客户进行不同处理，有效地提高了顾客满意率，成功再开发商业机会。
The customer management of huaxing changsha construction bank is a kind of marketing service manner in nature , which is a complete management of customer manager to customers and demonstrates features as the preliminary formation of categorized vip customers management system based on analysis of contributions to customer value . major customers will be guaranteed products and services of high quality , first priority , and encouraging favor so that good customer relationship can be preserved 长沙华兴建行的客户经理制实质上是一种营销服务方式，即客户经理对客户的全面管理，主要特点有：初步建立了以客户价值贡献分析法为分类标准的vip客户分类管理体系，对重点客户提供优质、优先、优惠的产品和服务，维护客户关系。
Outlier detection , an emergence of data mining relative to clustering , is to discovery the small scheme where the data have a notably different characteristic to other data objects . it can be used in fraud detection , e . g . , by detecting unusual usage of credit cards or telecommunications services 异常( outlier )检测则是数据挖掘中一个崭新的领域，用来发现“小的模式” (相对于聚类) ，即数据集中显著不同于其它数据的对象，异常检测在电信利信用卡欺骗、贷款审批、气象预报、金融领域和客户分类等各种应用。
