基于云模型的客户价值判定方法研究

发布时间：2018-05-19 20:21

本文选题：客户关系 + 云模型　；参考：《安徽农业大学》2013年硕士论文

【摘要】：当代社会企业快速发展，对客户资源的争夺日趋激烈。如何改善客户关系，获取客户评价，合理聚类客户，提高客户满意度是现代企业必须去深入思考、研究、投入财力人力的重要研究课题。基于评价模型的客户聚类方法研究，是当前该领域的研究热点，并金融财务，企业建设，客户管理等方面得到了广泛的应用。在评价模型的建模过程中，涉及到大量客户反馈的处理，这些知识的来源都是自然语言，往往都存在不确定性，因此，为了能构建出更加合理的、客观的评价模型，研究如何处理这些不确定性知识的理论与方法就显得十分必要。在此基础之上，对于不同满意度的客户如何聚类，按照何种指标以何种算法聚类也是本文所探讨的问题。针对在传统聚类以及评定方法失真或无效的情况下，本文较为深入的探讨了如何处理不确定性问题，重点如何解决客户评价建模中的关键问题——客户等级的云化问题。在客户价值等级模型建立的基础上，对各种客户聚类方法进行了深入的分析，设计并实现了一种使用云模型对客户评价指标聚类分析的方法。论文研究的主要内容及取得的成果如下： ①客户价值模型研究。通过处理海量的统计数据，对客户价值进行区分，并确定客户价值体系，利用MatlAB以及Sas进行分析，最后完成对客户聚类的工作。 ②提出了客户综合评价的云化方法。通过正向云化、逆向云化的方法从客户各个指标入手建立相应的评价模型，并进行主观模糊统计，最后将得到的统计数据通过云发生器得到不同客户的不同等级排名结果。 ③将聚类分析的结果与云化等级结果进行对比参照，互相验证，对企业客户的认识进一步深化，为企业决策提供依据和支持。论文研究成果对于客户评价建模理论与方法的进一步深入研究，，构建更加精确、更加客观的客户关系管理系统，进一步建立基于云模型客户聚类模型，实现客户价值的充分共享和协同服务，具有一定研究借鉴价值和实际应用意义。
[Abstract]:With the rapid development of modern social enterprises, the competition for customer resources is becoming increasingly fierce. How to improve customer relationship, obtain customer evaluation, reasonably cluster customers and improve customer satisfaction is an important research topic for modern enterprises to think deeply, study and invest in financial manpower. The research on customer clustering based on Evaluation Model is the current leader The research focus of the domain, and the financial finance, the enterprise construction, the customer management and so on, has been widely applied. In the modeling process of the evaluation model, it involves the treatment of a large number of customer feedback. The sources of these knowledge are natural language and often have uncertainty. Therefore, in order to build a more reasonable and objective evaluation model, It is very necessary to study the theory and method of dealing with these uncertain knowledge. On this basis, how to cluster the customers with different satisfaction and what kind of algorithm to cluster according to the index is also a problem discussed in this paper.
In the case of the distortion or ineffectiveness of traditional clustering and evaluation methods, this paper deeply discusses how to deal with the uncertainty problem and how to solve the key problem in customer evaluation modeling, the cloud problem of customer level. On the basis of the establishment of the customer value hierarchy model, various customer clustering methods are carried out. In depth analysis, a method of clustering analysis of customer evaluation indexes using cloud model is designed and implemented. The main contents and achievements of this paper are as follows:
(1) customer value model research. By processing massive statistical data, the customer value is distinguished, and the customer value system is determined. MatlAB and Sas are used to analyze the customer value. Finally, the customer clustering work is completed.
Secondly, a cloud based method of customer comprehensive evaluation is proposed. Through the forward cloud and reverse cloud method, the corresponding evaluation model is set up from each index of the customer, and the subjective fuzzy statistics are carried out. Finally, the results are obtained through the cloud generator to get the different ranking results of different customers.
Thirdly, the results of cluster analysis are compared with the results of cloud classification, mutual validation, further deepening of the understanding of enterprise customers, and providing the basis and support for enterprise decision-making. The research results of the paper further study the theory and methods of customer evaluation modeling, and build more accurate and more objective customer relationship management system. Further establish customer clustering model based on cloud model to achieve full value sharing and collaborative services, which has certain reference value and practical application significance.
【学位授予单位】：安徽农业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP311.13;F323.3

【参考文献】