Quantitive Analysis Of Sales And Credit Evaluation For C2C E-commerce Sellers Via Quantile Regression Approach

In recent years,C2 C e-commerce has rapidly developed.The turnover of C2 C ecommerce websites has improved at a blistering speed each year,such as Taobao.At the same time,the promotion of Internet and mobile and the perfection of virtual payment also have provided a high-speed channel for the development of C2 C e-commerce.More and more consumers have chosen online transaction.With the increasing online consumer groups,C2 C e-commerce has formed a new research area.However,during the process of the high-speed development,there came out some problems.Therefore,previous research has studied and researched for the field of C2 C e-commerce.Among them,the sales problems of electricity companies and credit evaluation have occupied much of the research work.And these works also have achieved good conclusions and results.But the existing literatures of C2 C e-commerce mainly used the mean regression analysis method that is difficult to consider the heterogeneity at the different quantiles.To this end,this dissertation uses quantile regression method for the research of sales problems and credit of C2 C electricity companies,and research the heterogeneity and law of fluctuation at different quantiles to obtain new discoveries.This dissertation carries out the empirical research in the following two aspects.(1)Analyzing the influencing factors of the sales in C2 C e-commerce platform.We choose the Taobao data that are grabbed as the research object and use the quantile regression neural networks(QRNN)to conduct the empirical research.The empirical research showed that lowest price,highest price,total number of comments and credit score have a heterogeneous effect to the sales.These factors have different effects to sales under the different levels of sales in degree and direction.(2)Discussing the mode of C2 C e-commerce sellers' credit scoring and proposing the method of validity test.Firstly,in order to solve the disadvantage of existing credit scoring mode,we establish a seller's credit scoring multifactorial correction model.Secondly,in order to test actual effect of method,we propose the concept of the cumulative consuming loss of shop to measure the loss that is caused by electricity companies to consumers and establish effective evaluation method of seller's credit score method based on QRNN model.Finally,By selecting the Taobao data that are grabbed as the research object and use the quantile regression neural networks(QRNN)to make the empirical research.The empirical research showed that the seller's credit scoring multifactorial correction model can effective reflect the real credit level since it can both successfully explain and accurately predict the cumulative loss of consumers.The research of the factors that influence the C2 C e-commerce sellers' sales can provide theoretical basis and reference for sellers,and help them to find an effective marketing strategy and increase sales.The improvement of C2 C e-commerce sellers' credit scoring mode can help sellers to evaluate sellers' credit level effectively,then that is helpful for sellers to improve the credit level of shop.