Design And Implementation Of Credit Risk Assessment System Based Social Network

With the rapid development of Internet technology,the development of Internet finance also presents an unprecedented prosperity.This results in a number of Internet financial models,among which Internet lending is one.However,due to the anonymity of Internet users,Internet users have high credit risk,among which the evaluation and prediction of personal credit risk of Internet users is the key to control credit risk of Internet lending platform.Research shows that social information has a significant impact on default risk,and the individual credit risk assessment model integrated with social information performs better.Therefore,reasonable and effective use of social information of network users will play a very important role in the assessment of personal credit risk.In view of this,this paper takes Sina Weibo as an example to analyze and study the basic methods of figure portraits.Afterwords this paper summarizes a set of construction methods of personas suitable for social network credit risk assessment.On this basis,this paper uses cart algorithm to build a certain number of decision trees.Then the random forest model is constructed.Afterwords the random forest model and logistic regression model are combined to realize the effective combination of machine learning algorithm and traditional credit evaluation method.The final model is named CreditPer model.Finally,this paper designs and implements a personal credit risk assessment system based on social network,which is divided into user collection module,user portrait building module,credit risk assessment module,data analysis module and system management module.These modules are responsible for the collection,construction,evaluation,analysis and management of users,so as to form a more complete credit risk assessment system.Experiments show that the combination model of CreditPer takes advantage of both the random forest model and the logistic regression model and improves the ability of differentiation to a large extent compared with the simple use of logistic model.At the same time,it also continues the explanation and stability of the traditional model.The research of this paper not only helps to promote the development of personal credit industry model and improves the control level of network loan risk in China,but also makes up for the deficiency of the third-party credit system in China.