The Optimization Of Credit Card Risk Management From Perspective Of Data Mining

With the development of economy,the reform of credit management system and the transformation and upgrading of consumption habits,China's credit card business has entered a stage of rapid development and has gradually become an important financial consumption tool.However,with the rapid growth of credit card market scale and business,illegal cash out,false use,overdue payment,malicious overdraft and other risk events continue to rise.Therefore,improving and strengthening the risk management of credit card business has become an important problem for commercial banks and banking regulatory departments.The insufficient information and asymmetric trading environment between banks and customers are the main factors leading to the occurrence of credit card risk.With the development of big data technology,commercial banks can obtain a large number of historical customer data,transaction data and behavior data more easily.Through the data mining technology to process and analyze the big data,we can get the relevant information and knowledge reflecting the user's credit,so as to help banks optimize the risk management scheme of credit card business.More and more scholars pay attention to the research fields of data mining technology and credit risk assessment,and create a variety of credit risk assessment models.However,through the analysis,it is found that most of the risk assessment models designed in the current literature focus on the pre event credit assessment stage,while the user behavior in the event and the collection behavior after the risk event are rarely involved.Therefore,from the perspective of information insufficiency and information asymmetry,this paper takes bank a as an example to study the risk management problems faced by the credit card business of commercial banks at this stage,and puts forward an active risk management optimization scheme based on data mining technology,including pre prevention,in-process monitoring and post control.This paper uses the research method of combining case analysis and normative analysis.First of all,it analyzes and summarizes the research status at home and abroad,condenses the key problems to be solved,clarifies the research theme and innovation points;second,it analyzes the business model,risk characteristics and management status of the credit card business,so as to lay a theoretical foundation for the research;third,it analyzes the data mining technology and its application in the credit card risk management to make itclear The feasibility and reliability of the application of the technical scheme.Finally,taking bank a as an example,aiming at the problems existing in the risk management of its credit card business,a specific optimization scheme based on data mining technology is proposed.The main conclusions are as follows:(1)The information asymmetry between banks and customers is the main factor leading to credit card risk.The main reasons for information asymmetry are: lack of perfect personal credit system,lack of legal system guarantee,lack of customer information database supported by big data,algorithm system,tracking and identification system and other control measures.(2)Under the background of the rapid development of mobile Internet,online and offline,on-site and network,time and space are rapidly integrated.It is difficult for traditional transaction monitoring system to find various risk events through established rules,and to defend against risk attacks in the new technology era.(3)At present,most banks' credit card risk management has some problems,such as single early warning dimension and transaction monitoring,backward risk management technology,and lack of perfect risk information sharing mechanism.(4)Data mining technology can mine valuable customer information hidden behind the data,establish risk factor identification and risk early warning model,and then reduce the risk probability brought by information asymmetry to the bank,change the traditional passive mode of risk management.