Research And Implementation Of Bank Account Risk Detection Based On Sequential Pattern Mining

With the rapid development of the modem economy and the continuous innovation of banks,E-banking has become one of the most important part in the banking business.However,in recent years,malfeasance towards electronic banking has driven a large number of users at risk,endangering the security of banks and users,and undermining social stability.The bank currently uses the rule and verification method to deal with such users at risk,which is inefficient and ineffective.How to find such risk users in a timely manner,whose information may be leaked and whose fiunds have been stolen in the account,become an important issue.In this paper,the related researches on bank user risk detection are summarized,and the theories of data mining and abnormal discovery are deeply studied.Based on the theoretical basis,this paper analyzes the characteristics of such risk users and proposes a method based on frequent sequence mining.The risk user detection model fir-stly performs fr-equent sequence mining on existing risk users,and then matches the mining result with risk user feature sequence fr-om the user which has been detected to have implement risk.The analysis of risk user bank flow data shows that,except for the characteristics of risk users with highly similar behavior sequences,the time interval has a great impact on the degree of user risk.Therefore,this paper takes the attribute of time interval as a key factor in the bank risk user detection model.Firstly,improve the frequent sequence mining algorithm with time interval by clustering the time interval.Second,adds time interval constraint to the traditional sequence pattern matching algorithm to improve Detection accuracy.According to the above algorithm,the proposed bank risk user detection model is designed and implemented in parallel with Hadoop environment.Finally,this paper conducts experiments and analysis on the proposed bank risk user detection model.The experimental results show that the risk user detection method described in this paper has certain validity and accuracy.It can effectively improve the accuracy and efficiency of bank risk user detection,so that banks can deal with risk users in time to ensure the safety of the account.At the same time,risk user information can be provided to related regulatory authorities to track and investigate criminal behaviors,and to maintain national security and social stability.