A Comprehensive Study of Data Mining-based Financial Fraud Detection Research

Financial Fraud Detection is a subject concerning many industries viz. banking, insurance, government agencies, law enforcement, Etcetera. A vast amount of financial exchanges and transactions happen daily and fraud cases are on a constant ascent. This makes the traditional methods of fraud detection very tedious to utilize and implement. Auditing the transactions manually is no longer feasible; and, therefore, such practices must constantly adapt to the increasing load. Use of data mining helps predict and, thus, quickly detect fraud, so that an organization can take immediate action to minimize effective costs. This study reviews the implementation and efficiency of Data Mining Methods- Decision Trees, Artificial Neural Networks, Logistic Model, and Bayesian Belief Networks, which offer fundamental solutions to predicaments in detecting Financial Frauds. This paper subsequently also addresses the challenges faced by the application of Data Mining Methods in Financial Fraud Detection (FFD) and the additional organisational requirements for its implementation.