AI-Driven Business Intelligence for Fraud Detection a Systematic Literature Review
Makaleler > Derleme · 2026
Yazarlar: İbrahim YÜLÜCE, Fatih Dagdeviren, Murat Orun, Ismail Duran, Orhan DAĞDEVİREN
2026 8th International Congress on Human-Computer Interaction, Optimization and Robotic Applications (ICHORA), Uluslararası, 2026, ss 1-9
DOI: 10.1109/ichora69329.2026.11537194
Özet
Due to the rapid growth of digital transactions, platform economies, and data-intensive business environments, financial fraud has become increasingly complex. In this context, the integration of artificial intelligence, business intelligence, and advanced analytics has emerged as a promising approach for improving fraud detection, prevention, and decision support systems. This systematic literature review aims to synthesize the existing evidence on how AI-driven analytical models and business intelligence environments are used together in financial fraud detection. The review was conducted in accordance with the PRISMA 2020 principles. Studies published from 2020 onwards were identified through major academic databases using predefined search strings. Following the removal of duplicates, title and abstract screening, and full-text eligibility assessment, 22 studies were included in the final synthesis. The findings indicate that the literature primarily focuses on General Financial Fraud, followed by Audit/Accounting Fraud and Transaction/Payment Fraud scenarios. In terms of operational timing, the majority of the examined solutions$(59 \%)$rely on offline processing, while the development of fully comprehensive real-time architectures remains limited. Furthermore, it has been determined that the biggest obstacles encountered in industrial integration are “scalability” and “data quality.” Converting risk scores derived from machine learning models into visual intelligence via interactive BI dashboards significantly increases the speed and accuracy of fraud detection by analysts. For technological investments to translate into operational value, organizations must design secure data pipelines and a data governance culture as complementary components.