Causal Temporal-Relation Contrastive Learning for Cryptocurrency Anti-Money Laundering: A Leakage-Aware Synthetic Benchmark

Authors

  • Celia Kumar Luddy School of Informatics, Computing, and Engineering, Indiana University Bloomington, USA
  • Elena Brooks College of Information Sciences and Technology, Pennsylvania State University, USA

DOI:

https://doi.org/10.66238/fsrma141

Keywords:

cryptocurrency AML, heterogeneous graph, contrastive learning, causal temporal evaluation, synthetic benchmark, label scarcity

Abstract

Cryptocurrency anti-money-laundering (AML) screening must account for heterogeneous transaction relations, evolving network activity, scarce labels and the information available when an alert is issued. Building on the relation-preserving contrastive design of HeteroGCL, this paper develops TA-HGCL, a causal temporal-relation encoder that combines typed historical neighborhood aggregation restricted to admissible predecessors, exponential recency weighting and feature-view contrastive regularization. Experiments on seeded Bitcoin-like transaction graphs with chronological training, validation and test partitions show that TA-HGCL attains the highest illicit-class F1 among the relation-based neural models (0.117 ± 0.058, versus 0.105 ± 0.049 for its causal no-contrastive variant and 0.092 ± 0.060 for its non-temporal variant). A future-feature perturbation audit confirms exact historical input invariance of the causal graph construction, and a feature-only logistic baseline (F1 0.180 ± 0.094) indicates that the simulated node attributes carry much of the predictive signal. The paper contributes an explicit decision-time information constraint, relation-specific temporal summaries and a reproducible evaluation protocol for leakage-aware AML research.

References

Foley, S., Karlsen, J. R., and Putniņš, T. J., “Sex, drugs, and bitcoin: How much illegal activity is financed through cryptocurrencies?” Review of Financial Studies, vol. 32, pp. 1798–1853, 2019. doi:10.1093/rfs/hhz015.

Weber, M., et al., “Anti-money laundering in bitcoin: Experimenting with graph convolutional networks for financial forensics,” arXiv:1908.02591, 2019.

Savage, D., Wang, Q., Chou, P., Zhang, X., and Yu, X., “Detection of money laundering groups using supervised learning in networks,” arXiv:1608.00708, 2016.

Colladon, A. F. and Remondi, E., “Using social network analysis to prevent money laundering,” Expert Systems with Applications, vol. 67, pp. 49–58, 2017.

Lorenz, J., Silva, M. I., Aparício, D., Ascensão, J. T., and Bizarro, P., “Machine learning methods to detect money laundering in the bitcoin blockchain in the presence of label scarcity,” Proc. ACM AI in Finance, 2020.

D. Teng, “TRIAD: Deadline- and reliability-aware scheduling for post-earthquake structural assessment workflows,” World Journal of Engineering Research, vol. 3, no. 6, pp. 55–66, 2025.

Hamilton, W. L., Ying, Z., and Leskovec, J., “Inductive representation learning on large graphs,” NeurIPS, 2017.

Veličković, P., et al., “Graph attention networks,” ICLR, 2018.

Wang, D., et al., “A semi-supervised graph attentive network for financial fraud detection,” IEEE ICDM, pp. 598–607, 2019.

Pareja, A., et al., “EvolveGCN: Evolving graph convolutional networks for dynamic graphs,” AAAI, vol. 34, pp. 5363–5370, 2020. doi:10.1609/aaai.v34i04.5984.

Xu, D., Ruan, C., Korpeoglu, E., Kumar, S., and Achan, K., “Inductive representation learning on temporal graphs,” ICLR, 2020.

Rossi, E., Chamberlain, B., Frasca, F., Eynard, D., Monti, F., and Bronstein, M., “Temporal graph networks for deep learning on dynamic graphs,” arXiv:2006.10637, 2020.

J. Chen, X. Zhang, and X. Han, “Explainable identification of tax-related financial reporting anomalies in U.S. public companies: Cross-fitted tax-consistency residuals and Owen-value Shapley propagation,” World Journal of Economics and Business Research, vol. 1, no. 1, pp. 33–45, 2023.

Hu, Z., Dong, Y., Wang, K., and Sun, Y., “Heterogeneous graph transformer,” WWW, pp. 2704–2710, 2020.

Chen, J., Liu, J., Liang, Y., and Zhou, M., “HeteroGCL: A heterogeneous graph contrastive learning framework for scalable and sustainable cryptocurrency AML,” Applied Sciences, vol. 16, no. 6, art. 2860, 2026. doi:10.3390/app16062860.

Dou, Y., et al., “Enhancing graph neural network-based fraud detectors against camouflaged fraudsters,” CIKM, 2020. doi:10.1145/3340531.3411903.

J. Chen, X. Han, X. Zhang, and Y. Yang, “Exact concept-level explanations for early detection of financial misstatements,” Eurasia Journal of Science and Technology, vol. 5, no. 2, pp. 34–48, 2023.

van den Oord, A., Li, Y., and Vinyals, O., “Representation learning with contrastive predictive coding,” arXiv:1807.03748, 2018.

Chen, T., Kornblith, S., Norouzi, M., and Hinton, G., “A simple framework for contrastive learning of visual representations,” ICML, pp. 1597–1607, 2020.

You, Y., et al., “Graph contrastive learning with augmentations,” NeurIPS, vol. 33, pp. 5812–5823, 2020.

Zhu, Y., Xu, Y., Yu, F., Liu, Q., Wu, S., and Wang, L., “Deep graph contrastive representation learning,” arXiv:2006.04131, 2020.

Hassani, K. and Khasahmadi, A. H., “Contrastive multi-view representation learning on graphs,” ICML, pp. 4116–4126, 2020.

Wang, X., Liu, N., Han, H., and Shi, C., “Self-supervised heterogeneous graph neural network with co-contrastive learning,” KDD, pp. 1726–1736, 2021.

X. Yue and R. Zhu, “An explainable hybrid machine learning framework for cash-flow-conditioned multi-horizon liquidity risk early warning: An SME-oriented benchmark study,” World Journal of Management Science, vol. 2, no. 1, pp. 63–72, 2024.

Saito, T. and Rehmsmeier, M., “The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets,” PLoS ONE, vol. 10, e0118432, 2015. doi:10.1371/journal.pone.0118432.

Y. Lin et al., “Subgraph-Level Temporal Graph Transformer With Laundering-Motif Prototypes for Anti-Money Laundering on Blockchains,” IEEE Access, 2026, doi: 10.1109/ACCESS.2026.3731160.

Y. Jiao, H. Fan, X. Yue, W. Ping, T. Sun, and J. Wang, “Dynamic heterogeneous graph contrastive learning for uncovering collusive financial fraud,” Scientific Reports, article in press, 2026, doi: 10.1038/s41598-026-58938-5.

Downloads

Published

2026-10-06