Supporting Decision-Making in Financial Investigations Through Prioritization of Suspicious Transactions Using Process Mining and Economic Evaluation دعم اتخاذ القرار في التحقيقات المالية من خلال ترتيب المعاملات المشبوهة حسب الأولوية باستخدام تحليل العمليات والتقييم الاقتصادي | ||||
منارة الاسكندرية للعلوم التجارية | ||||
Volume 1, Issue 2, July 2025, Page 147-116 PDF (1.13 MB) | ||||
Document Type: المقالة الأصلية | ||||
DOI: 10.21608/mauta.2025.388094.1001 | ||||
![]() | ||||
Authors | ||||
Mohamed S Abu-assi ![]() ![]() | ||||
المعهد المصري لأكاديمية الاسكندرية للإدارة والمحاسبة | ||||
Abstract | ||||
This research proposes a forensic approach based on process mining to detect suspicious transactions in financial event logs. Using real-world data from the BPI Challenge 2017, the study focuses on analyzing deviations in loan offer processes recorded by a Dutch financial institution. The approach begins with preprocessing and automated discovery of process models using algorithms such as Inductive Miner, followed by conformance checking via the Multi-Perspective Process Explorer to identify deviations from expected behavior. A degree of rarity of deviation degree is then introduced to prioritize anomalies, allowing investigators to focus on the least frequent and potentially fraudulent cases. A temporal analysis using dotted charts highlights delays and irregularities in specific activities. The findings reveal deviations not only in the timing but also in the roles of originators. To support financial investigations, the study includes economic interpretation of suspicious patterns, linking anomalies to potential financial impact and systemic risk. These findings suggest that integrating degree of rarity based process mining into financial investigations significantly improves anomaly detection and resource prioritization, making it a valuable addition to institutional forensic protocols. | ||||
Keywords | ||||
Process Mining; Financial Institutions; Economic Impact | ||||
Statistics Article View: 131 PDF Download: 26 |
||||