基于信息熵视角的在线零售企业顾客细分研究

发布时间：2018-01-05 04:35

本文关键词：基于信息熵视角的在线零售企业顾客细分研究　出处：《中国地质大学(北京)》2017年硕士论文　论文类型：学位论文

【摘要】：随着电子商务的繁荣发展,以及“互联网+”战略的推进,B2C这种商业模式越来越被广大企业和消费者所重视。与传统零售企业相比,在线零售企业的固定资产占比小、潜在顾客量大、产品服务不受时空限制等,使得顾客成为在线零售企业最重要的资产。CRM的有效实施为在线零售企业带来了全新的顾客关系管理方式,而对顾客进行科学细分是在线零售企业正确实施CRM的关键。并且,科学合理的顾客细分方法为在线零售企业实施差别化营销、精准营销等提供了基础,可有效增强企业竞争优势。因此,科学合理的顾客细分理论与方法研究具有重要的理论与现实意义。在线零售企业在顾客细分研究与实践中,在细分指标体系和模型两方面存在不足。在对顾客细分研究文献梳理的基础上,认为顾客细分的本质上是基于顾客属性和行为特征所包含的信息量展开的。基于此,将信息熵引入到顾客细分理论与方法研究中,构建基于RFM模型的顾客细分指标体系和基于信息熵和聚类分析方法相结合的顾客细分模型,最后对某在线零售企业一年的交易数据进行实证研究,基于顾客消费行为特征实现了顾客细分的目标,为在线零售企业的顾客细分实践提供理论与方法支持。具体而言,本文研究的主要成果包括:(1)将基于最大熵准则的客观赋权方法—熵权法应用到顾客细分的RFM模型的指标权重确定中,避免了指标权重主观设定的不足。(2)将叉熵用于对K-means聚类的改进中,通过熵值计算,基于样本之间的叉熵距离确定聚类数目和聚类中心,解决了经典K-means聚类方法中需要预先确定聚类数目和聚类中心的不足。(3)根据在线零售企业的特点,将基于叉熵的K-means聚类方法与改进的RFM模型相结合,构建了在线零售企业的顾客细分模型。并进行实证研究,分别从消费行为和顾客价值两个角度实现顾客的细分,通过更为细致的细分结果指导企业的经营决策,为企业实现精准营销提供决策支持。论文研究拓展了信息熵的应用领域,丰富了顾客细分的理论与方法,为在线零售企业的顾客细分提供了新的工具与方法,对于其他类似企业的顾客细分具有一定借鉴意义。
[Abstract]:With the development of electronic commerce, and to promote Internet plus "strategy, B2C business model is increasingly valued by the majority of enterprises and consumers. Compared with the traditional retail enterprise, fixed assets online retail enterprises accounted for a small amount of potential customers, products and services is not limited by time and space, so that customers become effective implementation online retail enterprises the most important asset of.CRM brings a new way of customer relationship management for online retail business, the customer segmentation is the key scientific and correct implementation of CRM online retail enterprises. And the scientific and reasonable method of customer segmentation for online retail enterprises to implement differentiated marketing, provides a basis for precision marketing, can effectively enhance the enterprise competitive advantage. Therefore, it has important theoretical and practical significance to research the customer segmentation theory and method of scientific and reasonable. The online retail enterprises in customer segmentation Research and practice, the problems existing in the two aspects of subdivision index system and model. Based on the customer segmentation research on the literature review, the essence of customer segmentation is the amount of information based on customer attributes and behaviors features expanded. Based on this, the research to the customer segmentation theory and method of information entropy is introduced. The construction of RFM model of customer segmentation index system and information entropy and cluster analysis method based on customer segmentation model based on the empirical research on a year of online retail transaction data, customer consumption behavior realizes customer segmentation based on the target of providing theory and method support for the practice of online customer segmentation of retail enterprises specifically, the main results of this study include: (1) the RFM model based on maximum entropy criterion of objective weighting methods - entropy method is applied to the customer segmentation. To determine the index weight, avoid the shortcoming of subjective index weight setting. (2) the cross entropy for the improvement of K-means clustering, through entropy calculation, determine the cluster number and center distance between the samples based on the cross entropy, solves the predetermined lack of cluster center and the number of the classic K-means clustering method. (3) according to the characteristics of online retail enterprises, the RFM model and the improved K-means clustering method based on the cross entropy combination, constructs the model of customer segmentation of online retail enterprises. And empirical research, separately from the two aspects of consumer behavior and customer value realization of customer segmentation, through a more detailed breakdown of the results to guide business decisions, provide decision support for enterprise to achieve precision marketing. This paper expands the application field of information entropy, enriches the theory and method of customer segmentation, for online retail enterprises The customer segmentation of the industry provides new tools and methods, which can be used for reference for the customer segmentation of other similar enterprises.

【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F274;F724.2

【相似文献】