Authors : G. Nandini; Jennifer Mary S.; Dr. Girish Kumar D.

Volume/Issue : Volume 11 - 2026, Issue 4 - April

Google Scholar : https://tinyurl.com/yc36e2zm

Scribd : https://tinyurl.com/2psyaebe

DOI : https://doi.org/10.38124/ijisrt/26apr1843

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : The rapid growth of digital payments, particularly through Card and UPI platforms, has increased exposure to fraud, unauthorized transactions, and behavioral anomalies. This paper presents CardSentinel, an adaptive fraud-detection system that integrates machine learning–based anomaly scoring, user behaviour analytics, geolocation verification, and device fingerprinting to assess transaction risk in real time. The system employs a rule-enhanced scoring engine combined with behavioural baselines—covering spending patterns, time-of-day activity, merchant familiarity, and location stability— to compute dynamic risk levels for every transaction. High-risk or unusual transactions trigger an OTP-based secondary verification layer to prevent unauthorized payments. The platform is developed using a modular microservice architecture with a React frontend, Node.js backend, and MySQL persistence layer, ensuring scalability, low latency, and secure data handling. Experimental evaluation across multiple usage scenarios demonstrates improved fraud-detection reliability, reduction of false alerts, and enhanced user safety. The paper discusses system architecture, module interactions, riskscoring logic, implementation strategies, and future enhancements involving predictive modeling and advanced behaviour profiling.

Keywords : Digital Payments, Fraud Detection, User Behaviour Analytics(UBA), UPI Security, Card Transaction Security, Machine Learning–Based Risk Scoring, Geolocation Verification, Device Fingerprinting, OTP Authentication, Node.js Backend, React Frontend, MySQL Database, Real-Time Fraud Analytics, Scalable Web Architecture.

References :

1. A. Kumar, R. Verma, and S. Nair, “Real-time fraud detection in digital payment systems using behavioural analytics,” IEEE Access, vol. 11, pp. 45231–45245, 2023.
2. P. Sharma and M. Gupta, “Adaptive risk scoring models for card and UPI transaction fraud prevention,” International Journal of Information Security, vol. 22, no. 4, pp. 615–628, 2023.
3. S. Rao, N. Iyer, and K. Mehta, “User behaviour profiling for financial fraud detection in online payment platforms,” Journal of Financial Crime, vol. 31, no. 1, pp. 89–104, 2024.
4. L. Chen and Y. Wang, “Device fingerprinting and user-agent analysis for securing online financial transactions,” IEEE Transactions on Information Forensics and Security, vol. 18, pp. 2174–2186, 2023.
5. R. Malhotra and A. Singh, “OTP-based authentication mechanisms for high-risk digital payment transactions,” IEEE Consumer Electronics Magazine, vol. 13, no. 2, pp. 48–56, 2024.
6. V. Patel, S. Shah, and R. Joshi, “Geolocation-aware anomaly detection for real-time payment fraud mitigation,” Computers & Security, vol. 132, pp. 103306, 2023.
7. T. Nguyen and H. Kim, “Design of scalable web-based dashboards for real-time fraud monitoring,” International Journal of Web Information Systems, vol. 20, no. 1, pp. 1–15, 2024.
8. A. Rahman and M. Siddiqui, “SMS and notification-based alert frameworks for secure financial systems,” IEEE Transactions on Human-Machine Systems, vol. 54, no. 1, pp. 62–71, 2024.
9. S. Banerjee and P. Das, “Secure backend architectures for financial transaction monitoring using Node.js and MySQL,” Journal of Cloud Computing, vol. 13, no. 1, pp. 1–14, 2024.
10. K. Zhou, L. Martin, and J. Brown, “A survey of machine learning techniques for digital payment fraud detection,” ACM Computing Surveys, vol. 57, no. 2, pp. 1–38, 2025.