基于聚类结果解释方法的客户群特征研究
|
摘要
市场经济的不断发展使得客户关系管理在企业经营管理中的重要性日益突出。客户细分作为客户关系管理的重要方法,可以将企业的客户进行有效划分,使企业能够根据不同的客户群特征制定相应的管理策略。在客户需求个性化和多样化的形势下,客户群特征的提取成为重要的研究课题。
针对细分后客户群特征难以提取问题,提出了基于聚类结果解释的特征提取方法。该方法将因子分析与聚类分析相结合,利用因子分析具有提取可解释因子的作用,对客户细分指标进行特征因子的提取,并计算客户聚类后各个客户群的因子得分的平均值,以表格和图形的方式简洁表示计算的结果。根据各个客户群特征因子值的大小及特征因子包含的客户指标对客户群特征进行分析说明。对方法的适用性及优越性进行了比较分析,说明了该方法在客户群特征研究中的合理性及科学性。
将提出的客户群特征分析方法应用于电信行业客户关系管理实证分析中,得到了最佳客户群及客户群特征因子值,对其进行分析不但能够对各个客户群特征清晰认识,解决了客户指标复杂化带来的客户群特征难以提取的问题;而且根据特征因子值的大小及特征因子包含的客户指标高度相关性原则,分析说明了各个客户群的特征,为企业制定交叉销售、捆绑销售以及开发新产品提供有价值的信息。相对以往的客户群特征提取和分析更有实用价值。
The importance of Customer Relationship Management (CRM) is growing in the company Management because of the accelerated development of the market economy. As the primary method of CRM, customer segmentation can help enterprises do customers division effectively, thus enterprises can make corresponding management strategy according to different customer group characteristics. In the situation of customer demand with personalization and diversification, the extraction of customer group characteristics has become an important research subject.
A method of feature extraction based on explanation of clustering results is proposed, due to the problem of having difficulty in extracting customer group characteristics after subdivision. Combined with factor analysis and cluster analysis and taking advantage of one point that factor analysis can extract explained factors, this method can extract characteristic factors from customer segmentation indexes, and calculate the average factor score of each customer group after customer clustering, then show calculation results by tables and charts. In this way, customer group characteristics can be analysed and interpreted in the light of different factor value in each customer group characteristics and customer indexes included in characteristic factors. After comparing and analysing its applicability and superiority,the method is proved reasonable and scientific in the study of the customer group characteristics.
When the method is applied to empirical analysis of telecom customer relationship management, we can get optimum customer groups and their characteristics' factor value. Through analysing the results, we can not only know better each customer group characteristics, solving the problem of having difficulty in extracting customer group characteristics that is brought about by the complication of customer indexes; but also analyse each customer group characteristics according to the factor value and information contained in factors highly correlated with customer indexes, providing valuable information for enterprises to make cross-selling, bundling selling and develop new products. Compared with the previous methods, this method doing customer group characteristics' extraction and analysis has more practical value.
针对细分后客户群特征难以提取问题,提出了基于聚类结果解释的特征提取方法。该方法将因子分析与聚类分析相结合,利用因子分析具有提取可解释因子的作用,对客户细分指标进行特征因子的提取,并计算客户聚类后各个客户群的因子得分的平均值,以表格和图形的方式简洁表示计算的结果。根据各个客户群特征因子值的大小及特征因子包含的客户指标对客户群特征进行分析说明。对方法的适用性及优越性进行了比较分析,说明了该方法在客户群特征研究中的合理性及科学性。
将提出的客户群特征分析方法应用于电信行业客户关系管理实证分析中,得到了最佳客户群及客户群特征因子值,对其进行分析不但能够对各个客户群特征清晰认识,解决了客户指标复杂化带来的客户群特征难以提取的问题;而且根据特征因子值的大小及特征因子包含的客户指标高度相关性原则,分析说明了各个客户群的特征,为企业制定交叉销售、捆绑销售以及开发新产品提供有价值的信息。相对以往的客户群特征提取和分析更有实用价值。
The importance of Customer Relationship Management (CRM) is growing in the company Management because of the accelerated development of the market economy. As the primary method of CRM, customer segmentation can help enterprises do customers division effectively, thus enterprises can make corresponding management strategy according to different customer group characteristics. In the situation of customer demand with personalization and diversification, the extraction of customer group characteristics has become an important research subject.
A method of feature extraction based on explanation of clustering results is proposed, due to the problem of having difficulty in extracting customer group characteristics after subdivision. Combined with factor analysis and cluster analysis and taking advantage of one point that factor analysis can extract explained factors, this method can extract characteristic factors from customer segmentation indexes, and calculate the average factor score of each customer group after customer clustering, then show calculation results by tables and charts. In this way, customer group characteristics can be analysed and interpreted in the light of different factor value in each customer group characteristics and customer indexes included in characteristic factors. After comparing and analysing its applicability and superiority,the method is proved reasonable and scientific in the study of the customer group characteristics.
When the method is applied to empirical analysis of telecom customer relationship management, we can get optimum customer groups and their characteristics' factor value. Through analysing the results, we can not only know better each customer group characteristics, solving the problem of having difficulty in extracting customer group characteristics that is brought about by the complication of customer indexes; but also analyse each customer group characteristics according to the factor value and information contained in factors highly correlated with customer indexes, providing valuable information for enterprises to make cross-selling, bundling selling and develop new products. Compared with the previous methods, this method doing customer group characteristics' extraction and analysis has more practical value.
引文
[1] Anton J. The past, present and future of customer access centers[J]. International Journal of Service Industry Management, 2000,11(2):120-130.
[2] Lee D. What is Customer Relationship Management[J]. European Management Journal, 2000,8(3):102-113.
[3] Peppers D, Rogers M, etc. Is your Company Ready for one to one Marketing[J]. Harvard Business Review, 1999(12):151-160.
[4]杨琴. CRM营销时代[J].中外管理,2000,10(8):69-71.
[5] Kaj S. Segmentation Based on Customer Profitability-Respective Analysis of Retail Bank[J]. Journal of Marketing Management, 1997,5(13):479-492.
[6] Berry M J, Linolf G. Data Mining Technique of Marketing, Sales and Customer Support[J]. Marketing Research, 1997,3(6):35-42.
[7] Peppard J. Customer Relationship Management in Financial Services[J]. European Management Journal, 2000,18(3):312-327.
[8] Bolton P. Dynamic A. Model of the Duration of the Customer’s Relationship with a Continuous Sevrice Provider: the Role of Satisfaction[J]. Marketing Science, 1998,17(l):45-65.
[9] Bore P, Johnson L. A Customer-Service Work Relationship Model[J]. International Journal of Service Industry Management, 2000,5(11):491-511.
[10] Needll B, Robert L. How to Lose Your Star Performers without Losing Customers[J]. Harvard Business Review, 2001,79(10):104-112.
[11] Ha S, Park S. Customer’s Time-Variant Purchase Behavior and Corresponding Marketing Strategies: an Online Retailer’s Case[J]. Computers & Industrial Engineering, 2006,43(4):801-820.
[12] Kamakura W A, Wedel W. Amazon Cross-Selling through Database Marketing: a Mixed Data Factor Analyzer for Data Augmentation and Prediction[J]. International Journal of Research in Marketing, 2006,20(l):45-65.
[13] Verhoef F, Hoekstra J C. The Impact of Satisfaction and Payment Equity on Cross-Buying: A Dynamic Model for a Multi-Service Provider[J]. Journey of Retailing, 2007,77(3):359-378.
[14]李怀祖,奇佳音.客户关系管理(CRM)的体系框架分析[J].工业工程,2002,5(1):42-45.
[15]张喆，?鹑唬菩≡?数据挖掘技术在CRM中的应用[J].中国管理科学,2003,11(1):53-56.
[16]胡丽,张卫国,康治平.保险业CRM系统实现的关键技术研究[J].科技进步与对策,2006,23(9):143-145.
[17]胡少东.分析型客户关系管理(CRM)在移动电话行业的应用[J].价值工程,2006,25(8):134-137.
[18]唐敏,刘红平.信息技术在CRM中的应用研究[J].西南农业大学(自然科学版),2006,28(3):503-506.
[19]吴晓东,陈纯.基于策略的CRM决策支持研究[J].模式识别与人工智能,2007,20(6):861-866.
[20]谷再秋,于福.客户关系管理(CRM)系统功能分析[J].中国管理信息化,2009,12(12):71-73.
[21]郭迎春.知识型客户关系管理研究[J].生产力研究,2009(2):141-143.
[22]张俊光.项目客户关系管理实施方法研究[J].生产力研究,2009(18):72-74.
[23]方芳,刘大有,王新.电信CRM技术发展研究[J].计算机工程,2010,36(5):277-281.
[24]夏勇,金卫健.基于工作流程技术的客户关系管理系统研究[J].中国管理信息化,2010,13(3):88-91.
[25] Smith W. Product Differentiation and Market Segmentation. Alternative Marketing Strategies[J]. Journal of Marketing, 1996(7):3-8.
[26] Morwitz V G, Schmittlein D. Using Segmentation to Improve Sales Forecasts Based on Purchase Intent: Which“Intenders”Actually Buy[J]. Journal of Marketing Research, 1992(29):391-405.
[27] Balakrishnan P V, Cooper M C, Jacob V S. Comparative Performance of the FSCL Neural Net and K-means Algorithm for Market Segmentation[J]. European Journal of Operational Research, 1996(93):346-357.
[28] Kuo R J, Ho L M, Hu C M. Cluster Analysis in Industrial Market Segmentation through Artificial Neural Network[J]. Computers and Industrial Engineering, 2002,42(2):391-399.
[29] Hwang H, Jung T, Suh E. Model and Customer Segmentation Based on Customer Value: A Case Study on the Wireless Telecommunication Industry[J]. Expert System with Applications, 2005(26):181-182.
[30] Dalton P. Customer Segmentation cues to Support Market Segmentation[J]. Journal of Marketing Research, 2008,14(9):11-13.
[31]郭蕴华,陈定方.基于模糊聚类分析的客户分类算法研究[J].计算机应用研究,2005(4):52-53.
[32]周颖,吕巍,井淼.基于数据挖掘技术的移动通信行业客户细分[J].上海交通大学学报,2007,41(7):1142-1145.
[33]邵良杉,王鹤.蚁群聚类组合算法在证券行业客户细分中的应用[J].计算机系统应用,2008(3):84-86.
[34]向坚持,刘相滨,资武成.基于密度的K-means算法及在客户细分中的应用研究[J].计算机工程及应用,2008,44(35):246-248.
[35]张东玲,高齐圣,李朝玲.基于二元语义的客户群聚类分析方法[J].工业工程与管理,2008,13(3):46-51.
[36]钱维佳,王延清.基于动态聚类的证券业客户细分实证研究[J].计算机应用,2010,30(2):495-498.
[37]彭昂,王如龙,陈泉泉.基于复杂属性相似度的聚类算法及其应用研究[J].计算机应用,2010,30(7):1930-1932.
[38]时红军,韩兵.改进K-means算法在B2C电子商务客户细分中的应用[J].微型电脑应用,2010,26(10):39-42.
[39]吴春旭,鲍满园,荀清龙.自组织映射聚类算法在电信客户细分中的应用[J].计算机系统应用,2010,19(8):168-171.
[40]吴斌,郑毅,傅伟鹏.一种基于群体智能的客户行为分析算法[J].计算机学报,2003,26(8):913-918.
[41]路晓伟,蒋馥,侯立文.基于客户本体的客户特征提取[J].计算机工程,2005,31(5):31-33.
[42]卢致杰.基于距离的客户群价值研究[J].管理工程学报,2006,20(1):18-23.
[43]李凯,田双亮,耿丽君,张喜.基于改进遗传算法在分析企业客户群中的应用[J].河南理工大学学报(自然科学版),2009,28(6):811-815.
[44] Burghard C, Galimi J. Customer Relationship Management[M]. Gartner Advisory, 2000:12-13.
[45] Dorigo M, Bonabeau E, Theraulaz G. Ant algorithms and stigmergy[J]. Future Generation Computer System, 2000(6):851-871.
[46]王卫平,杨杰.基于蚁群智能的客户群偏好分析方法[J].管理科学,2005,18(4):53-57.
[47]任永功,于戈.一种多维数据的聚类算法及其可视化研究[J].计算机学报,2005,28(11):1861-1865.
[48]刘子维,孟波,王超.一种聚类挖掘结果的可视化方法[J].计算机应用研究,2006,23(5):75-76.
[49]许翔燕,江永全,杨燕,张仕斌.聚类结果可视化研究[J].微计算机信息,2007,23(12):90-91.
[50]熊才权,李德华,张玉.研讨厅专家意见聚类分析及其可视化[J].模式识别与人工智能,2009,22(2):282-287.
[51]陈英,顾国昌,吕天阳.基于离群点识别的聚类结果属性特征簇发现[J].哈尔滨工程大学学报,2009,30(3):312-317.
[52] Rigby D K, Reichheld F F, Schefter P. Avoid the Four Perils of CRM[J]. Harvard Business Review, 2002(l):101-109.
[53]陈治平,胡宇舟,顾学道.聚类算法在电信客户细分中的应用研究[J].计算机应用,2007,27(10):2566-2577.
[54]冯晓蒲,张铁峰.四种聚类方法比较[J].微型机与应用,2010,29(16):1-3.
[55]范明,戴冠中,覃森.基于有效性函数的聚类算法[J].计算机科学,2007,34(5):197-200.
[56] Hand D, Marmila H, Smyth P.数据挖掘原理[M].张银奎,廖丽,宋俊,译.北京:机械工业出版社,2003:179-180.
[57] Hand D, Mannila H, Smyth P. Principles of Data Mining[M]. Cambridge: The MIT Press, 2001:131-142
[58] Julio J V. Interpreting fuzzy clustering results with virtual reality-based visual data mining: application to microarray gene expression data[J]. North American Fuzzy Information Processing Society (NAFIPS), 2004(6):27-30.
[2] Lee D. What is Customer Relationship Management[J]. European Management Journal, 2000,8(3):102-113.
[3] Peppers D, Rogers M, etc. Is your Company Ready for one to one Marketing[J]. Harvard Business Review, 1999(12):151-160.
[4]杨琴. CRM营销时代[J].中外管理,2000,10(8):69-71.
[5] Kaj S. Segmentation Based on Customer Profitability-Respective Analysis of Retail Bank[J]. Journal of Marketing Management, 1997,5(13):479-492.
[6] Berry M J, Linolf G. Data Mining Technique of Marketing, Sales and Customer Support[J]. Marketing Research, 1997,3(6):35-42.
[7] Peppard J. Customer Relationship Management in Financial Services[J]. European Management Journal, 2000,18(3):312-327.
[8] Bolton P. Dynamic A. Model of the Duration of the Customer’s Relationship with a Continuous Sevrice Provider: the Role of Satisfaction[J]. Marketing Science, 1998,17(l):45-65.
[9] Bore P, Johnson L. A Customer-Service Work Relationship Model[J]. International Journal of Service Industry Management, 2000,5(11):491-511.
[10] Needll B, Robert L. How to Lose Your Star Performers without Losing Customers[J]. Harvard Business Review, 2001,79(10):104-112.
[11] Ha S, Park S. Customer’s Time-Variant Purchase Behavior and Corresponding Marketing Strategies: an Online Retailer’s Case[J]. Computers & Industrial Engineering, 2006,43(4):801-820.
[12] Kamakura W A, Wedel W. Amazon Cross-Selling through Database Marketing: a Mixed Data Factor Analyzer for Data Augmentation and Prediction[J]. International Journal of Research in Marketing, 2006,20(l):45-65.
[13] Verhoef F, Hoekstra J C. The Impact of Satisfaction and Payment Equity on Cross-Buying: A Dynamic Model for a Multi-Service Provider[J]. Journey of Retailing, 2007,77(3):359-378.
[14]李怀祖,奇佳音.客户关系管理(CRM)的体系框架分析[J].工业工程,2002,5(1):42-45.
[15]张喆，?鹑唬菩≡?数据挖掘技术在CRM中的应用[J].中国管理科学,2003,11(1):53-56.
[16]胡丽,张卫国,康治平.保险业CRM系统实现的关键技术研究[J].科技进步与对策,2006,23(9):143-145.
[17]胡少东.分析型客户关系管理(CRM)在移动电话行业的应用[J].价值工程,2006,25(8):134-137.
[18]唐敏,刘红平.信息技术在CRM中的应用研究[J].西南农业大学(自然科学版),2006,28(3):503-506.
[19]吴晓东,陈纯.基于策略的CRM决策支持研究[J].模式识别与人工智能,2007,20(6):861-866.
[20]谷再秋,于福.客户关系管理(CRM)系统功能分析[J].中国管理信息化,2009,12(12):71-73.
[21]郭迎春.知识型客户关系管理研究[J].生产力研究,2009(2):141-143.
[22]张俊光.项目客户关系管理实施方法研究[J].生产力研究,2009(18):72-74.
[23]方芳,刘大有,王新.电信CRM技术发展研究[J].计算机工程,2010,36(5):277-281.
[24]夏勇,金卫健.基于工作流程技术的客户关系管理系统研究[J].中国管理信息化,2010,13(3):88-91.
[25] Smith W. Product Differentiation and Market Segmentation. Alternative Marketing Strategies[J]. Journal of Marketing, 1996(7):3-8.
[26] Morwitz V G, Schmittlein D. Using Segmentation to Improve Sales Forecasts Based on Purchase Intent: Which“Intenders”Actually Buy[J]. Journal of Marketing Research, 1992(29):391-405.
[27] Balakrishnan P V, Cooper M C, Jacob V S. Comparative Performance of the FSCL Neural Net and K-means Algorithm for Market Segmentation[J]. European Journal of Operational Research, 1996(93):346-357.
[28] Kuo R J, Ho L M, Hu C M. Cluster Analysis in Industrial Market Segmentation through Artificial Neural Network[J]. Computers and Industrial Engineering, 2002,42(2):391-399.
[29] Hwang H, Jung T, Suh E. Model and Customer Segmentation Based on Customer Value: A Case Study on the Wireless Telecommunication Industry[J]. Expert System with Applications, 2005(26):181-182.
[30] Dalton P. Customer Segmentation cues to Support Market Segmentation[J]. Journal of Marketing Research, 2008,14(9):11-13.
[31]郭蕴华,陈定方.基于模糊聚类分析的客户分类算法研究[J].计算机应用研究,2005(4):52-53.
[32]周颖,吕巍,井淼.基于数据挖掘技术的移动通信行业客户细分[J].上海交通大学学报,2007,41(7):1142-1145.
[33]邵良杉,王鹤.蚁群聚类组合算法在证券行业客户细分中的应用[J].计算机系统应用,2008(3):84-86.
[34]向坚持,刘相滨,资武成.基于密度的K-means算法及在客户细分中的应用研究[J].计算机工程及应用,2008,44(35):246-248.
[35]张东玲,高齐圣,李朝玲.基于二元语义的客户群聚类分析方法[J].工业工程与管理,2008,13(3):46-51.
[36]钱维佳,王延清.基于动态聚类的证券业客户细分实证研究[J].计算机应用,2010,30(2):495-498.
[37]彭昂,王如龙,陈泉泉.基于复杂属性相似度的聚类算法及其应用研究[J].计算机应用,2010,30(7):1930-1932.
[38]时红军,韩兵.改进K-means算法在B2C电子商务客户细分中的应用[J].微型电脑应用,2010,26(10):39-42.
[39]吴春旭,鲍满园,荀清龙.自组织映射聚类算法在电信客户细分中的应用[J].计算机系统应用,2010,19(8):168-171.
[40]吴斌,郑毅,傅伟鹏.一种基于群体智能的客户行为分析算法[J].计算机学报,2003,26(8):913-918.
[41]路晓伟,蒋馥,侯立文.基于客户本体的客户特征提取[J].计算机工程,2005,31(5):31-33.
[42]卢致杰.基于距离的客户群价值研究[J].管理工程学报,2006,20(1):18-23.
[43]李凯,田双亮,耿丽君,张喜.基于改进遗传算法在分析企业客户群中的应用[J].河南理工大学学报(自然科学版),2009,28(6):811-815.
[44] Burghard C, Galimi J. Customer Relationship Management[M]. Gartner Advisory, 2000:12-13.
[45] Dorigo M, Bonabeau E, Theraulaz G. Ant algorithms and stigmergy[J]. Future Generation Computer System, 2000(6):851-871.
[46]王卫平,杨杰.基于蚁群智能的客户群偏好分析方法[J].管理科学,2005,18(4):53-57.
[47]任永功,于戈.一种多维数据的聚类算法及其可视化研究[J].计算机学报,2005,28(11):1861-1865.
[48]刘子维,孟波,王超.一种聚类挖掘结果的可视化方法[J].计算机应用研究,2006,23(5):75-76.
[49]许翔燕,江永全,杨燕,张仕斌.聚类结果可视化研究[J].微计算机信息,2007,23(12):90-91.
[50]熊才权,李德华,张玉.研讨厅专家意见聚类分析及其可视化[J].模式识别与人工智能,2009,22(2):282-287.
[51]陈英,顾国昌,吕天阳.基于离群点识别的聚类结果属性特征簇发现[J].哈尔滨工程大学学报,2009,30(3):312-317.
[52] Rigby D K, Reichheld F F, Schefter P. Avoid the Four Perils of CRM[J]. Harvard Business Review, 2002(l):101-109.
[53]陈治平,胡宇舟,顾学道.聚类算法在电信客户细分中的应用研究[J].计算机应用,2007,27(10):2566-2577.
[54]冯晓蒲,张铁峰.四种聚类方法比较[J].微型机与应用,2010,29(16):1-3.
[55]范明,戴冠中,覃森.基于有效性函数的聚类算法[J].计算机科学,2007,34(5):197-200.
[56] Hand D, Marmila H, Smyth P.数据挖掘原理[M].张银奎,廖丽,宋俊,译.北京:机械工业出版社,2003:179-180.
[57] Hand D, Mannila H, Smyth P. Principles of Data Mining[M]. Cambridge: The MIT Press, 2001:131-142
[58] Julio J V. Interpreting fuzzy clustering results with virtual reality-based visual data mining: application to microarray gene expression data[J]. North American Fuzzy Information Processing Society (NAFIPS), 2004(6):27-30.