Research On C2C E-Commerce Business Credit Evaluation Model

In recent years,with the rise of the mobile Internet,the number of Internet users in China has rapidly increased,making e-commerce gradually gaining recognition from people and becoming a part of people's lives.Compared with traditional transactions,e-commerce is more convenient,transparent,and has more options.The interaction between businesses and buyers is stronger.Multi-data shows that the growth of ecommerce in China is still considerable.However,with the rapid growth of ecommerce,some new problems have also hindered the development of e-commerce.The openness and virtuality of the Internet have brought new fertile ground for some fraudulent activities.New types of fraud have emerged in an endless stream.The C2 C e-commerce features have made it possible.As the “hard-hit areas” and “high-risk areas” for these fraud problems,the fraud of C2 C e-commerce transactions has received more and more attention.The lack of integrity has brought great harm to the normal ecommerce transaction order,increased transaction costs,and brought a severe test to the development of e-commerce in China.This will not only hinder the development of e-commerce in China,but also disrupt the normal Market Order.Although major e-commerce platforms have established their own evaluation mechanisms for these issues,although they have effectively curbed frauds at the beginning of their establishment,it is easy for unscrupulous merchants to use their loopholes to implement fraud.Based on a large number of researches on existing results,this article uses Taobao as an example to summarize the main problems existing in the e-commerce seller credit evaluation model,and to address these issues from the improvement of the e-commerce credit evaluation index to the establishment of a model.these questions.This article adopts the experimental method and according to the results obtained by the Department of Network Supervision of the State Administration for Industry and Commerce in 2015 on online shopping goods sampling and reporting — the authenticity rate of Taobao.com is 37.25%.We will simulate the proportion of the business people in accordance with the ratio of 37.25% and 62.75%.Divided into two groups,a group of businesses that simulate good faith,another group of businesses that simulate fraud,and the process of trading with buyers simulating different credit levels,to obtain the basic data needed for model training,and to improve the evaluation indicators: As for the transaction amount,this article considers the degree of deviation of the commodity price from the average price of the commodity as a consideration;for the buyer's credit,this paper adopts different weights for evaluations made by buyers with different credit levels.For the issue of time,According to the distance from the current time,the assessment will be given different weights.The return of positive feedback will be included in the credit evaluation index system.The consideration of price will be added.The goods with high prices will have a greater weight in the role of credit,and the commodity price will be the same as the average commodity in the industry.The price is compared to see whether it deviates from a reasonable range and combines the existing descriptions of Taobao.The existing indicators of logistics,logistics,and services improve the indicator system.In the aspect of model establishment: This paper analyzes the advantages and disadvantages of existing data mining algorithms,uses the naive Bayes algorithm and BP neural network in data mining to design the dynamic model of C2 C e-commerce credit evaluation,and uses the naive Bayes classifier to do preliminary screening.Then use the BP neural network to make further judgments on the integrity of the merchant to get the final result.Finally,the obtained model is compared with the existing model using only one method to verify the accuracy of the model.