MONEY LAUNDERING DETECTION USING GRAPH NEURAL NETWORKS ENHANCED WITH AUTOENCODER COMPONENTS
DOI:
https://doi.org/10.24193/subbi.2025.06Keywords:
money laundering, graph neural networks, autoencoders, explainabilityAbstract
The paper addresses the topic of detecting money laundering operations in transaction data represented as graph data-structures. We propose the integration of autoencoder components in Graph Neural Networks (GNN) architectures, in order to incorporate a reconstruction step in the traditional edge classification problem and enhance model quality based upon the usage of reconstruction errors.
We show that enhancing GNNs with autoencoder components improves the predictive performance of money laundering detection, on data represented as homogeneous graphs. Additionally, the Shapley value is computed in order to gain further insight into the most important features from distinguishing normal and fraudulent activities.
2010 Mathematics Subject Classification. 68T05, 68T99.
1998 CR Categories and Descriptors. I.2.6 [Artificial Intelligence]: Learning – Connectionism and neural nets; E.1 [Data]: Data structures – Graphs and networks.
References
[1] Abal, R., Pen˜a, L., and Soria Quijaite, J. Machine learning models for money laundering detection in financial institutions. a systematic literature review. In Proceedings of the 22nd LACCEI International Multi-Conference for Engineering, Education, and Technology (2024), pp. 1–10.
[2] Altman, E. R., Egressy, B., et al. Realistic Synthetic Financial Transactions for Anti-Money Laundering Models. In Proceedings of NeurIPS 2023 (2023), pp. 1–24.
[3] Battaglia, P. W., Hamrick, J. B., et al. Relational inductive biases, deep learning, and graph networks. CoRR abs/1806.01261 (2018), 1–40.
[4] Cassara, J. A. Countering international money laundering: Total failure is ”only a decimal point away”. Tech. rep., The FACT Coalition, Washington, DC, August 2017.
[5] Egressy, B., von Niederha¨usern, L., et al. Provably Powerful GNNs for Directed Multigraphs. In Proceedings of (AAAI-24) (2024), AAAI Press, pp. 11838–11846.
[6] Gilmer, J., Schoenholz, S. S., Riley, P. F., Vinyals, O., and Dahl, G. E. Neural message passing for quantum chemistry. CoRR abs/1704.01212 (2017), 1–14.
[7] Hu, W., Liu, B., Gomes, J., Zitnik, M., Liang, P., Pande, V. S., and Leskovec, J. Pre-training graph neural networks. CoRR abs/1905.12265 (2019), 1–22.
[8] Jaume, G., Nguyen, A., et al. edGNN: a Simple and Powerful GNN for Directed Labeled Graphs. CoRR abs/1904.08745 (2019), 1–9.
[9] Jensen, R. I. T., and Iosifidis, A. Fighting money laundering with statistics and machine learning. IEEE Access 11 (2023), 8889–8903.
[10] Johannessen, F., and Jullum, M. Finding money launderers using heterogeneous graph neural networks. CoRR abs/2307.13499 (2023), 1–20.
[11] Kipf, T. N., and Welling, M. Variational graph auto-encoders. CoRR abs/1611.07308 (2016), 1–3.
[12] Kumar, S., Ahmed, R., et al. Exploitation of machine learning algorithms for detecting financial crimes based on customers’ behavior. Sustainability 14, 21 (2022), 1–24.
[13] Lundberg, S. M., Erion, G., et al. From local explanations to global understanding with explainable ai for trees. Nature Machine Intelligence 2, 1 (2020), 2522–5839.
[14] Lundberg, S. M., and Lee, S.-I. A Unified Approach to Interpreting Model Predictions. In Adv Neural Inf Process Syst 30. Curran Associates, Inc., 2017, pp. 4765–4774.
[15] Murphy, A., Robu, K., and Steinert, M. The investigator-centered approach to financial crime: Doing what matters. McKinsey and Company (June 2020). Accessed: March 8, 2025.
[16] Richardson, D., Williams, and Mikkelsen, D. Network analytics and the fight against money laundering. McKinsey and Company, 2019.
[17] Sato, R., Yamada, M., and Kashima, H. Approximation ratios of graph neural networks for combinatorial problems. CoRR abs/1905.10261 (2019), 1–15.
[18] Schlichtkrull, M. S., Kipf, T. N., et al. Modeling Relational Data with Graph Convolutional Networks. In Proceedings of ESWC 2018, (2018), vol. 10843 of Lecture Notes in Computer Science, Springer, pp. 593–607.
[19] Suzumura, T., and Kanezashi, H. Anti-Money Laundering Datasets: InPlusLab anti-money laundering datadatasets. http://github.com/IBM/AMLSim/, 2021.
[20] Tang, M., Yang, C., and Li, P. Graph auto-encoder via neighborhood wasserstein reconstruction. CoRR abs/2202.09025 (2022), 1–17.
[21] Velickovic, P., Fedus, W., Hamilton, W. L., Lio`, P., Bengio, Y., and Hjelm, R. D. Deep graph infomax. CoRR abs/1809.10341 (2018).
[22] Verhage, A. Supply and demand: anti-money laundering by the compliance industry. Journal of Money Laundering Control 12 (10 2009), 371–391.
[23] Weber, M., Domeniconi, G., Chen, J., Weidele, D. K. I., Bellei, C., Robinson, T., and Leiserson, C. E. Anti-money laundering in bitcoin: Experimenting with graph convolutional networks for financial forensics. CoRR abs/1908.02591 (2019), 1–7.
[24] Xu, K., Hu, W., Leskovec, J., and Jegelka, S. How powerful are graph neural networks? CoRR abs/1810.00826 (2018), 1–17.
[25] You, J., Gomes-Selman, J. M., Ying, R., and Leskovec, J. Identity-aware graph neural networks. Proceedings of the AAAI Conference on Artificial Intelligence 35, 12 (May 2021), 10737–10745.
[26] Zhou, J., Cui, G., Zhang, Z., Yang, C., Liu, Z., and Sun, M. Graph neural networks: A review of methods and applications. CoRR abs/1812.08434 (2018), 57–81.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Studia Universitatis Babeș-Bolyai Informatica
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
