Detection Of Identity Theft In Credit Card Application Forms Through Data Mining Techniques Utilizing Multilayer Algorithms

Abstract

Due to the unpredictable nature of the market, the deceleration in economic growth, and the swift surge of digital e-commerce, the issue of fraud has gained extensive prevalence. As electronic commerce technology continues to rapidly evolve, credit card usage has witnessed a surge, establishing it as the preferred payment method for both online and offline transactions. Consequently, the escalation of credit card fraud has emerged, impacting customers seeking smart cards and loans, who now have the option to apply for credit cards through online channels or traditional paper forms. Regrettably, these application processes have brought to light occurrences of fraudulent activities, notably identity theft, posing a critical concern for both credit cardholders and financial institutions. Ill-intentioned individuals are illicitly acquiring customers' identities and gaining unauthorized access to credit cards, thereby exposing substantial risks for both customers and financial entities.

Nonetheless, the current strategies reliant on business rules and scorecards for fraud detection, excluding data mining, have exhibited shortcomings. In response to these limitations, this research introduces an innovative approach for real-time fraud detection during the application phase. This approach entails the implementation of a novel multi-layer fraud tracking system founded on data mining algorithms. This system incorporates two distinct algorithms, namely communal tracing and spike tracing, synergizing to enhance the precision, swiftness, and efficiency of fraud detection procedures. By validating applications in real-time upon submission, this system serves as a robust deterrent against the approval of fraudulent credit card applications prior to issuance.