Research Of Network Consumer Value Based On Data Mining

With the increasing number of online shopping,the number of e-commerce enterprises is increasing,the competition between enterprises is becoming more and more fierce,to retain and attract valuable consumers is the key to enhance the competitiveness of enterprises.The current research about the value of consumer is mostly based on the customer's perspective.There are few researches from the perspective of the enterprise.And most of the research perspective is single,it only focus on the factors of consumer value when making the classification,without considering the factors of consumer behavior.Therefore,through comparing the research contents and methods about related consumer value at home and abroad,and combining with the literature,the network consumer value model based on RFCP model is proposed and the comprehensive analysis model about network consumer value which combined with clustering and decision tree algorithm is established,then the empirical analysis is carried out.The specific research contents are as follows:Firstly,the data should be prepared and the indicators should be determined.The data are selected,which includes the consumer's own characteristic,such as product information and transaction data between consumers and enterprises.Four indicators are selected for the new model about consumer value calculation,the time interval which from recent purchase time to the present time,the frequency of purchase,the costs of purchase and the profits.The indicators used to classify and predict the value of the network consumers,such as the type of production,order source.Secondly,based on the traditional RFM model the improved RFCP model is proposed.The network consumers is divided into three groups,to achieve the similarity maximization of consumer value within the group and achieve the similarity minimization of consumer value between groups,which combines with clustering analysis method.The DEA(Data Envelopment Analysis)cross efficiency evaluation method is used to calculate the network consumer value,and three groups network consumers of with different value are sorted,then the different decision tree algorithms are used to classify and predict those network consumers with different values.Again,the online sales data of Jingdong mall are selected.And the empirical analysis is carried out on the consumer value model,which classifies the different value consumers for the enterprise,and forecasts the new customer.Thus,the practicability of the comprehensive analysis model of network consumer value is verified.Finally,the consumer value hierarchy strategy is analyzed.The necessity of constructing differentiated services and the marketing strategies for different value network consumers are proposed.Enterprise designs targeted marketing strategies focus on different value of network consumers,then the utilization rate of resources is improved.Enterprise's resource allocation receives optimization,it not only provides high quality service for consumers,also can enhance the market share and competitiveness and meet the needs of sustainable development,which has certain social significance.