《计算机应用论文面向汽车营销的客户管理模型与算法的设计与实现.doc》由会员分享，可在线阅读，更多相关《计算机应用论文面向汽车营销的客户管理模型与算法的设计与实现.doc（5页珍藏版）》请在三一办公上搜索。
1、面向汽车营销的客户管理模型与算法的设计与实现 面向汽车营销的客户管理模型与算法的设计与实现 Design and Implementation of Customer Management Model and Algorithm in Automobile Marketing-Oriented 【摘要】 中国在加入WTO之后,其汽车市场成为各大汽车公司争夺的焦点,越来越多的国外汽车制造企业(如大众、通用、福特、丰田、现代等)进入中国,再加上本土的一些民营汽车企业以及国有的汽车企业,中国的汽车市场呈现了百花争鸣的现象。08年共有50多个新车型上市,今年又将有60多个车型推出,中国汽车销售竞争越演
2、越烈,而且影响汽车销售的因素众多,特别客户作为企业的重要资源,他们需求往往都不一样,受地区、文化等影响较大。因此如何在最短的时间内挖掘出潜在客户,明确客户的价值和满意度,采取有效的措施防范客户流失,从而准确分析客户的需求,降低营销成本,预测市场走向,将成为汽车制造企业生存发展的关键因素之一。本文作为面向汽车营销的智能决策支持系统(国家863计划)的重要组成部分,是在客户关系管理(CRM)相关理论的指导下,运用商务智能(BI)、数据挖掘(DM)等相关技术,对汽车营销部门的客户基础信息、车辆信息、购买信息、服务信息等进行了分析。具体包括:使用数据挖掘经典聚类算法K-means,根据客户的价值对客户
3、进行细分,分析同一类别中的客户具有的相似属性;使用数据挖掘经典关联规则算法Apriori对潜在客户进行识别,找出购买某一车型的客户具有的共性;使用数据挖掘经典分类方法ID3决策树分析客户的流失问题,找出导致客户流失的具体原因,指定相应的客户保持策略;最后使用模糊综合评判方法对客户的满意度进行综合分析,找出薄弱环节,最终将满意度上升为忠诚度。最后,对以上工作进行了总结,并指出下一步的研究方向。【Abstract】 Chinas automobile industry after the initial stage of development, is now entered a fast per
4、iod, production and sales in recent years have maintained a high economic growth. Chinas vehicle production is the fourth world ranking currently, and is almost the same with Germany who ranked third. The result of rising vehicle production is the inevitable competition, in 2008 the profits of major
5、 domestic automakers decreased substantially, auto dealers were got into a vicious inevitable price competition, and now productions “blowout” and prices “avalanche “comment on the mass media are a common vocabulary. In this context, the major automobile manufacturers have increased the investment i
6、n informationization, such as Supply Chain Management (SCM) systems, Enterprise Resource Planning (ERP) systems, Customer Relationship Management (CRM) systems and so on.Automotive manufacturing and marketing business after many years of informatization construction, information systems have had mas
7、s production and operation data, but all of them have not been fully utilized yet, so the next focus of informationization construction in automotive industry is to carry out the construction of decision support functions, which required IT system provide more functions to support decision-making, a
8、nd gradually phase from general business stage to Business Intelligence stage, providing technical support to the automotive manufacturing enterprises from the production and operation type to the operation decision-making type.At the same time, an enterprise to survive and develop,it must know the
9、market, know customers and establish a “market-oriented, customer-centric” business philosophy, auto dealers as a services sector particular should be. Customer Relationship Management (CRM) is a client-driven, customer-centric decision-making technology which is in order to maintain the competitive
10、 power. To achieve these, it is necessary to collection, collation and analysis the variety of customer data produced in the process of customers interact with enterprise, then we can mine useful information and knowledge in these data. At present, bring data mining technology into the customer rela
11、tionship management is the key to improve the efficiency of decision-making, it can extract useful information with a large amount of data, predict trends in customer behavior, better support business decisions.Data mining, is the process of extracting knowledge which people interested from the data
12、base. Such knowledge is implicit, in advance unknown, potentially useful information, knowledge are represented to concept, rules, laws, models and other forms. Data Mining technologies in CRM application provide a strategic and decision-making support in the development of enterprises, contribute t
13、o find business development trend of automobile enterprises, predict unknown results, and help auto dealers analysis the key factor in necessary tasks, in order to achieve the purposes of increasing revenues, lowering costs and being an advantageous competitive position.Customer segmentation refers
14、to dividing enterprises existing customers into different customer clusters in accordance with the definite standard. The right customer segmentation can effectively reduce costs, while a stronger, more profitable market penetration. Customer segmentation allows marketing and sales personnel, as wel
15、l as decision-making platform to observe the client information from a relatively high level, allows companies to use the appropriate marketing strategy with different types of customers, and improves objectives and effectiveness of enterprise marketing services activities, thereby reducing marketin
16、g costs relatively, developing and maintaining the client resources maximally. Such as using clustering algorithm of data mining to cluster customer basic information, vehicle purchase information, customer service information in auto dealer company, from the substantial surface unrelated client inf
17、ormation we can find useful information to auto dealers, such as what characteristics the high-value customers have, what are their shopping habits, what background information and so on. then the company could provide adequate technical and human resources support to these high value customers to m
18、eet their demands and expectations, thereby maximizing the profits of enterprises.Potential customer identification. For most enterprises, the development of new customers is a new way of profit, but also is the main way of business growth. In this process, we must first clear the characteristics of
19、 different customers. That is, where the target market? What are the potential customers? What are the potential customers are high-quality clients? What is the degree of difficulty to obtain customers? All of these are the work of customers identification. Second, for different clients adopt differ
20、ent marketing strategies. Finally, according to reflect situation of the customer on marketing efforts, adjust the target customers and the corresponding marketing strategies. Identification of potential customer is the basic step to the other two steps, and it is a very important guiding role. In a
21、utomobile-oriented marketing decision support system, using the history sales data and related customers basic data in the same car model, and taking advantage of association rules algorithm in data mining, to identify the common characteristics of these clients, and their similar patterns of behavi
22、or, and then identify potential customers who will have a positive response, all of these can help auto dealers implement targeted marketing campaigns, improve work efficiency.Customer loss and maintenance. With the increasingly fierce competition in automotive industry, to obtain a new customers sp
23、ending become more and more, and keep customers cost savings than to acquire new customers, so to maintain their original customers is becoming increasingly valuable. Improvement of the way to maintain customers is to take action before the really loss of customers, this is the value of the loss mod
24、el. Automobile enterprises can use data mining techniques, such as decision tree classification algorithm, analysis the mass customer data in customer database, investigate the lost client groups, so as to analyze their characteristics, set up the lost customers model, and identify the patterns whic
25、h cause the loss of customers. Then according to the results of the analysis to identify the possible lost customers, and combine with the loss client model, analysis the data mining result, we can predict which customers might leave, so that the car sales company can formulate a number of appropria
26、te plans and programs to induce potential lost customers to stay, so as to providing personalized service, achieving “one-on-one” marketing, improving the relationship between company and customers, striving to maintain client and enhancing revenue.Customer satisfaction analysis. Customer satisfacti
27、on is the satisfaction when customer face to enterprise, its products and services, it is a subjective feel of the customer, it is also customers emotional performance to the products and services, or to the enterprises. The fundamental aim of enhancing satisfaction is to reduce the wastage rate of
28、customers, to enhance loyalty, and ultimately to increase the profits of enterprises. With sustained development of the automobile market in China, the level of customer satisfaction becomes an important aspect together with other indicators in measuring the competitiveness of enterprises. First of
29、all, high customer satisfaction will increase the purchase frequency and volume of the companys products or services; Secondly, high satisfaction customers have more loyalty and higher customer retention rate; Finally, high satisfaction customers will be more willing to recommend the companys produc
30、ts or services to others and bring new customers. The use of fuzzy comprehensive evaluation method, analysis satisfaction survey score in terms of price, quality, brands, service of the same auto model, and then arrive at a comprehensive evaluation of the model, as well as the number of share ratio
31、at different evaluation criteria in the four evaluation components. All of these can intuitive point out the factors which affect the satisfaction of this model, auto dealers can utilize the results to do targeted improvement, thereby enhancing the customers satisfaction and loyalty of the products
32、and services. 【关键词】 商务智能(BI); 决策支持系统; 客户关系管理(CRM); 数据挖掘(DM) 【Key words】 Business Intelligence; Decision Support System; Customer Relationship Management; Data Mining 面向汽车营销的客户管理模型与算法的设计与实现内容提要 4-7 第1章 绪论 7-15 1.1 研究背景 7-10 1.1.1 汽车营销现状及其决策支持系统的发展 7-8 1.1.2 客户关系管理在汽车营销中的应用 8-10 1.2 面向汽车营销的智能决策支持系统 10
33、-13 1.2.1 系统体系结构 10-11 1.2.2 数据的收集与整合 11-12 1.2.3 使用的软件及开发环境 12-13 1.3 本文的主要工作 13-15 第2章 核心技术概述 15-21 2.1 商务智能 15-17 2.1.1 决策支持系统与商务智能的关系 15 2.1.2 商务智能的层次和发展趋势 15-17 2.2 数据挖掘 17-19 2.2.1 数据挖掘模式 17-18 2.2.2 数据挖掘过程 18-19 2.3 客户关系管理（CRM） 19-21 2.3.1 CRM 的目的 19-20 2.3.2 CRM 的作用 20-21 第3章 客户价值细分 21-28 3.
34、1 客户细分问题 21-22 3.1.1 客户细分背景 21 3.1.2 客户价值细分 21-22 3.2 客户价值细分过程 22-26 3.2.1 K-means 聚类算法 22-24 3.2.2 客户价值细分模型 24-26 3.3 客户价值细分结果 26-28 第4章 潜在客户识别 28-34 4.1 潜在客户识别问题 28-29 4.1.1 潜在客户识别的任务 28 4.1.2 潜在客户识别的方法 28-29 4.2 潜在客户识别过程 29-33 4.2.1 Apriori 关联规则算法 29-30 4.2.2 识别潜在客户模型 30-33 4.3 潜在客户识别结果 33-34 第5章
35、 客户流失与保持 34-43 5.1 客户流失分析问题 34-35 5.1.1 客户流失现状和危害 34-35 5.1.2 客户流失原因和分类 35 5.2 客户流失分析过程 35-40 5.2.1 决策树、ID3 算法 36-38 5.2.2 客户流失分析模型 38-40 5.3 客户流失分析结果 40-41 5.4 客户保持 41-43 第6章 客户满意度分析 43-49 6.1 客户满意度分析问题 43-44 6.1.1 客户满意度的形成过程 43-44 6.1.2 客户满意度的研究进展 44 6.2 客户满意度分析过程 44-46 6.2.1 模糊综合评判法 44-45 6.2.2 客户满意度分析模型 45-46 6.3 客户满意度分析结果 46-47 6.4 从客户满意到客户忠诚 47-49 第7章 总 结 49-50 7.1 论文主要工作概述 49 7.2 下一步工作 49-50 参考文献 50-52 致谢 52-53 摘要 53-56 Abstract 56-59
