Research On E-Commerce Precision Marketing Data Mining Method

Technology's advancement has greatly promoted the development of social production,and the process of e-commerce has gradually accelerated.Social life has undergone tremendous changes,business contacts have become more frequent,customers have actively participated in social production,whose personalized and autonomous needs have brought more development space for e-commerce development.At the same time,competition and opportunities coexist,and the degree of convenience is accompanied by an increase in customer selectivity.E-commerce companies must grasp the opportunities and generate revenues in a changing development environment,and understand customers in a more multidimensional way.The way to grasp and cherish the individualized needs of different customers and implement precise marketing strategies is necessary.This is inextricably linked to the key factors that focus on analyzing customer value.Based on the customer knowledge,the mining and analysis of the potential consumer demand can help the company to find more influential factors of potential benefits,and lay the foundation for fully realizing the double maximization of the company's own resources and customer satisfaction.Based on the previous studies,this paper focuses on the implementation strategies and characteristics of each field of precision marketing.Selecting the fast-moving goods ecommerce customers as the research object,based on the classic marketing theories such as4 P,4C and customer transfer value,the analysis of the influencing factors of customer value of e-commerce enterprise is carried out in terms of product,price,promotion,consumer,cost and convenience.Firstly,this paper defines the customer value as the customer's contribution to the enterprise.Secondly,on the basis of the research on the influencing factors of the customer's value,the customer's objective habits are considered,including the customer's purchase time habits and receiving habits.By analyzing the relationship between customer purchase feature attributes,it is found that the influencing factors of customer value are complex and changeable.For this reason,the decision tree model with high degree of nonlinear fitting is considered for analysis.Through simulation and comparative analysis,this paper improves from two aspects,one is by means of Bagging sampling,the other is to use Boosting gradient descent method to carry out experiments,and by comparing and comprehensive analysis,the error gradient rate of the model is finally determined.On this basis,through the comparative analysis before and after,it is verified that the objective habits of customers in both aspects play an important role in customer value prediction,and the importance of the customer's purchase of feature attributes is output.Thirdly,it analyzes the correlation of important attribute characteristics of customers,and compares the characteristics of customer purchase attributes in different value intervals,which lays a foundation for e-commerce companies to specify differentiated marketing strategies for different customer value groups.