Abdul Azeem, Mohammed and Tanvir Rahman, Akash and Ismoth, Zerine (2022) BUSINESS RULES AUTOMATION THROUGH ARTIFICIAL INTELLIGENCE: IMPLICATIONS ANALYSIS AND DESIGN. International Journal of Economy and Innovation, 29. pp. 381-404. ISSN 2545-0573

Text 381-404+BUSINESS+RULES+AUTOMATION+THROUGH+ARTIFICIAL+INTELLIGENCE.pdf Download (888kB)

Abstract

BRMS have long been the structural backbone of organizational decision-making, and support the definition, validation and enforcement of operational logic in compliance, financial transactions and fraud prevention areas. However, the classic inflexible rule-based approaches can have problems in keeping up with a fast-growing information load and changing tactics of attackers. The study aims to explore the revolutionary nature of Artificial Intelligence (AI) in the automated process of business rules, and its implication to systems analysis and design. Based on the Credit Card Transactions Fraud Detection Dataset (2019-2020), the study shows how AI-based models, which comprise decision trees, ensemble learning, and explainable AI, can be used to derive adaptive rules based on transactional data. The study uses a simulated dataset of 1,000 customers and 800 merchants to investigate the trends of fraudulent and honest activity on each type of transaction, demographic, geography, and time. Such categories as grocery point of sale, online shopping, and food and dining are determined as the most vulnerable and have gender peculiarities, with female customers having higher rates of fraud affecting personal care product purchases and male customers having higher fraud rates affecting online shopping. Geographic analysis demonstrates that there are high fraud concentrations in states with high populations like California, Texas, Florida, and New York, whereas the temporal analysis indicates that these periods have high consumption. The paper also deals with the ethical issues such as privacy of the data, fairness, and transparency of the data, which will make AI usage responsible. All in all, by combining AI-based models and automated business rules, precision, scalability, and efficiency in fraud detection are improved and can help to design intelligent, adaptive, and ethically sound financial security systems.

Item Type:	Article
Subjects:	H Social Sciences > HJ Public Finance
Divisions:	Postgraduate > Master's of Islamic Education
Depositing User:	Journal Editor
Date Deposited:	07 Oct 2025 11:53
Last Modified:	07 Oct 2025 11:53
URI:	http://eprints.umsida.ac.id/id/eprint/16415

Actions (login required)

View Item