Fraud Detection by Monitoring Customer Behavior and Activities

Abstract

With the enhancement in technology e-banking like credit Card, Debit Card, Mobile Banking and Internet Banking is the popular medium to transfer the money from one account to another. E-Banking is gaining popularity day by day, which increases the online transaction with the increase in online shopping, online bill payment like electricity, Insurance Premium and other charges, online recharges and online reservation of railways, bus etc., so the fraud cases related to this are also increasing and it puts a great burden on the economy, affecting both customers and financial bodies. It not only costs money, but also a great amount of time to restore the harm done. The purpose is to prevent the customer from online transaction by using specific technique i.e. based on Data Mining and Artificial Intelligence technique. The risk score is calculated by Bayesian Learning Approach to analyze whether the transaction is genuine or fraudulent based on the two parameters: Customer Spending Behaviour and Geographical Locations. The customer than spending behaviour that can be identified by KMEAN clustering algorithm and in geographical location the current geographical location is compared with the previous location. If risk score is greater 0.5 then transaction is considered to be fraudulent transactions and then the security mechanism authenticates the user by entering the 4 digit random number that appears on the screen and the genuine user enters the code in a correct manner.