Ali, Mohammad and Hossain, Md Shahadat and Haider, Professor Parvin Sultana (2025) Explainable AI-Driven Behavioral Analytics for Detecting Emerging Financial Crime Patterns in Digital Banking Systems. American Journal of Technology Advancement, 2 (12). ISSN 2997-9382

Text AJTA_4220_Explainable+AI-Driven+Behavioral.pdf Download (610kB)

Abstract

With the rise in digital banking and online payment systems, financial crimes have become more complex and sophisticated, necessitating the adoption of intelligent and transparent fraud detection methods. The study aims to explore how Explainable Artificial Intelligence (XAI) through Behavioral Analytics can be leveraged for the detection of emerging patterns in financial crime within digital banking systems. The study uses an exploratory research approach with the secondary research methodology. R Studio is used to conduct exploratory analysis of the chosen dataset to understand transaction patterns, risk factors, device patterns, and payment processes related to fraudulent transactions. We conducted a secondary research analysis of existing scholarly work on AI, behavioral analytics, machine learning fraud detection, and explainable AI. The outcomes of visualization analysis are important patterns in relation to fraud distribution, abnormal transaction behavior, behavioral risk indicators, and device based fraud patterns. The results prove the effectiveness of the detection of financial crimes based on several behavioral characteristics and not one. The study demonstrates that the combination of explainable AI and behavioral analytics can enhance the transparency, accountability, and understanding of AI fraud detection decisions. The findings of this study will be useful in the development of responsible AI solutions to enhance digital banking security and aid in the identification of new or changing financial crime schemes.

Item Type:	Article
Uncontrolled Keywords:	Digital Banking, Online Payment, Financial crimes, Behavioral Analytics, Fraudulent Transaction, Explainable Artificial Intelligence, R Studio, AI, Dataset, Exploratory Study
Subjects:	H Social Sciences
Depositing User:	admin eprints
Date Deposited:	02 Jul 2026 04:09
Last Modified:	02 Jul 2026 04:09
URI:	http://eprints.umsida.ac.id/id/eprint/16747

Actions (login required)

View Item