基于数据挖掘的网络消费者价值研究

发布时间：2018-01-05 11:39

本文关键词：基于数据挖掘的网络消费者价值研究　出处：《哈尔滨商业大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着网购人数的不断增多,电子商务企业的数量在不断增加,企业之间竞争也越来越激烈,而保留和吸引有价值的消费者是企业提升竞争力的关键。当前消费者价值研究大部分是基于客户的视角,很少站在企业的视角去研究;并且大部分研究在对客户进行分类时只考虑到消费者价值因素,没有考虑到消费者行为因素,研究角度单一。因此,通过对比分析国内外有关消费者价值的研究内容及方法,结合文献,提出了基于RFCP模型的网络消费者价值评价指标选定,同时结合聚类和决策树算法建立了网络消费者价值综合分析模型,并进行了实证分析。具体研究内容如下:首先,数据的准备与指标的确定。选择消费者自身特征数据,如:产品信息及消费者与企业间的交易行为等相关数据。选择消费者最近一次购买时间到实验日期的时间间隔、购买频次、购买产品的成本、购买产品给企业带来的利润4个指标作为新消费者价值计算模型指标。产品种类、订单来源等指标用于消费者价值分类及预测。其次,在传统RFM模型的基础上提出了改进的RFCP模型。采用聚类分析法将网络消费者划分成三组,实现组内消费者价值相似度最大化,组间消费者价值相似度最小化;再结合DEA(Data Envelopment Analysis数据包络分析)交叉效率评价法来计算每组网络消费者价值,可以对三组不同价值的网络消费者进行排序;之后采用决策树算法对不同价值的网络消费者进行分类及预测。再次,选取京东商城的在线销售数据,对消费者价值模型进行实证分析,为企业分类不同价值消费者,并且对新增的消费者进行分类预测,从而验证了网络消费者价值综合分析模型的实用性。最后,消费者价值层次策略分析。提出构建差异化服务的必要性以及企业针对不同价值网络消费者的营销对策。企业针对不同价值的消费者制定针对性的营销策略,提高资源利用率,企业的资源配置得到优化的同时不仅可以为消费者提供优质的服务,还可以提升企业的市场占有率和竞争力,满足可持续发展的需要,具有一定的社会意义。
[Abstract]:With the increasing number of online shopping, the number of e-commerce enterprises are increasing, competition is becoming more and more intense among enterprises, and to retain and attract valuable consumers is the key to enhance the competitiveness of enterprises. The research value of consumer is mostly based on the perspective of the customer, rarely stand in the perspective of research; and most of the research in the classification of customers when considering only the consumer value factors, without considering the factors of consumer behavior, single research angle. Therefore, through the study of comparative analysis of domestic and foreign consumers value the contents and methods, combined with the literature, the selected evaluation indexes of network consumer value based on the RFCP model, combined with clustering and decision tree algorithm is constructed. Comprehensive analysis of online consumer value model, and empirical analysis. The specific contents are as follows: first, the preparation of data and means The subject of consumer choice determined. Characteristics of data, such as data and product information between consumers and business transactions. A recent consumer choice to buy time to date of the experiment time interval, purchase frequency, purchase cost, the purchase of products to bring the profits of enterprises 4 indicators as the new consumer value calculation model index. The product category, order source index for consumer value classification and prediction. Secondly, based on the traditional RFM model and improved RFCP model is proposed. By using the cluster analysis method of network consumers are divided into three groups, the group of consumers within the value maximizing the similarity between groups, consumer value similarity minimization; combined with the DEA (analysis of Data Envelopment Analysis) data envelopment evaluation method to calculate the efficiency of each cross network consumer value, of three groups of different value of network consumption By sorting; after using decision tree algorithm for different values of network consumer classification and prediction. Thirdly, online sales data from the Jingdong of the mall, the empirical analysis of consumer value model for enterprise value classification of different consumers, and to classify new consumers, which verified the practicality of the network consumer value analysis model. Finally, the analysis of consumer value level strategy. Put forward the necessity of differentiated services and enterprises for different marketing strategy of value network consumers. According to different enterprise value of consumers to develop targeted marketing strategies, improve the utilization rate of resources, enterprises optimize the allocation of resources at the same time not only can provide quality services consumers can also enhance the market share and competitiveness, to meet the needs of sustainable development, with a The social significance of setting.

【学位授予单位】：哈尔滨商业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F724.6;F274

【参考文献】