Methodology for updating the knowledge base of the fraud monitoring system

The severity of the problem of the growth of financial fraud forces researchers and practitioners to constantly consider options for the development of fraud monitoring systems. However, despite all technological efforts to resolve the problem, the volume of stolen funds is growing annually. This suggests a direct connection between new types of fraud and a negative trend in the growth of damage. In this regard, this article is aimed at identifying the possibility of replenishing a set of known fraud in the fraud monitoring system through synthetic, but potentially possible scenarios. The article discusses the processes of preparing data for training machine learning models and a complete algorithm for the interaction of classifying and generative models. Random Forest was chosen as the classifying model, and CTGAN was chosen to generate realistic data. The link for the models is the initialization of a JSON file with metadata containing the quality metrics of the Random Forest model and a list of popular attributes of fraud operations. As a result of the experiment, it was possible to create realistic examples of fraud operations based on combinations of fraud features with the highest impact factor. The results reveal the potential for automating the process of replenishing internal fraud databases, the information from which serves as the basis for setting up fraud identification systems.

1. Babanskaya A.S., Ermoleva D.R., Efimenko N.A., Akulova S.A. Benchmarking and opportunities of AI technologies for fraud prevention in the financial sector. Economics. Information Technologies. 2025;52(1):110–124. (In Russ.). https://doi.org/10.52575/2687-0932-2025-52-1-110-124

2. Larionova S.L., Ryakhovski E.E. Improvement of anti-fraud system algorithms based on the use of Graph Representation Learning methods and CycleGAN. Innovation & Investment. 2021;(6):137–142. (In Russ.).

3. Ali A., Abd Razak Sh., Othman S.H., et al. Financial fraud detection based on machine learning: a systematic literature review. Applied Sciences. 2022;12(19):9637. https://doi.org/10.3390/app12199637

4. George Z.H., Alam Kh., Hasan T. Machine learning for fraud detection in digital banking: a systematic literature review. ASRC Procedia: Global Perspectives in Science and Scholarship. 2023;3(1):37–61. https://doi.org/10.63125/913KSY63

5. Boubker M.B., Eddaoui A., Ouahabi S., Guemmat K.E., Chafik T. A systematic review of credit card fraud detection and prevention techniques. Nanotechnology Perceptions. 2024;20(S14):471–500. https://doi.org/10.62441/nano-ntp.vi.2802

6. Yussiff A.-S., Prikutse L.F., Asuah G., et al. The best machine learning model for fraud detection on e-platforms: a systematic literature review. Computer Science and Information Technologies. 2024;5(2):195–204. https://doi.org/10.11591/csit.v5i2.pp195-204

7. Khabibrahmanov R., Panteleeva L. Application of machine learning classification methods to detect fraudulent transactions. Vestnik Universiteta upravleniya "TISBI". 2025;(4):88–102. (In Russ.).

8. Mathpati Sh.Sh., Kaladeep Yalagi P.C., Athavale V.A., Azarov I. A Comparative Analysis of Machine Learning Models for Financial Fraud Detection. In: AISMA-2025: International Workshop on Advanced Information Security Management and Applications, 11–15 May 2025, Stavropol, Russia. Cham: Springer; 2026. P. 323–334. https://doi.org/10.1007/978-3-032-07275-7_30

9. Serzhan Ye., Umarov T. Fraud detection in credit card transactions using machine learning: a comparative analysis. Universum: Technical Sciences. 2025;(5-9). https://doi.org/10.32743/UniTech.2025.134.5.20106

10. Krajnov N.M., Bobrov L.K. "White hacker" as an option for using artificial intelligence in anti-fraud systems. In: Enterprise Engineering and Knowledge Management: Collection of Scientific Papers of the XXVII Russian Scientific Conference, 28–29 November 2024, Moscow, Russia. Moscow: Plekhanov Russian University of Economics; 2024. P. 168–171. (In Russ.).