GRAFraud: Leakage-Safe Graph-Augmented Multimodal Financial Fraud Detection with Temporal Corporate Relational Evidence

Authors

  • Jisoo Kang Korea University Business School, Seoul, Republic of Korea

DOI:

https://doi.org/10.66421/int.j.datasci.intell.syst64

Keywords:

financial fraud detection, heterogeneous graph, multimodal learning, corporate relations, temporal leakage, imbalanced classification

Abstract

Financial fraud screening benefits from evidence that is dispersed across accounting variables, narrative disclosure, and corporate relationships. This paper presents GRAFraud, a leakage-safe graph-augmented multimodal framework that combines numerical financial indicators and report narratives with temporally valid relational evidence derived from auditor, director, and repeated-firm links. The method first produces out-of-sample numerical and textual risk scores under a rolling chronology. It then constructs a heterogeneous corporate graph and extracts two classes of relational evidence: smoothed historical enforcement exposure from prior years and label-free same-year peer-risk propagation. A relation-aware meta-classifier is combined with a graph-free multimodal score stack through a validation-selected evidence gate. The resulting design preserves a transparent evidence path from each prediction to its numerical, textual, and relational contributors. Experiments use a controlled synthetic benchmark containing 8,400 firm-year observations for 1,400 firms over 2015–2020, with 459 fraud cases (5.46%), 40 auditors, 535 directors, 12 industries, 14 numerical indicators, and narrative disclosures. A strict temporal protocol trains base models on 2015–2018, tunes on 2019, and tests on 2020; hidden synthetic variables are excluded. On the 2020 test set, GRAFraud reaches AUC-ROC 0.846, average precision 0.332, F1 0.374, precision 0.377, recall 0.372, and MCC 0.338. Relative to raw early multimodal fusion, the graph-augmented model improves F1 by 8.19 points and MCC by 7.24 points while also increasing AUC and average precision. The results support temporally constrained relational evidence as a practical complement to filing-level multimodal signals.

 

References

[1] M. D. Beneish, 'The Detection of Earnings Manipulation,' Financial Analysts Journal, vol. 55, no. 5, pp. 24-36, 1999.

[2] P. M. Dechow, W. Ge, C. R. Larson, and R. G. Sloan, 'Predicting Material Accounting Misstatements,' Contemporary Accounting Research, vol. 28, no. 1, pp. 17-82, 2011.

[3] F. Li, 'Annual report readability, current earnings, and earnings persistence,' Journal of Accounting and Economics, vol. 45, no. 2-3, pp. 221-247, 2008.

[4] T. Loughran and B. McDonald, 'When Is a Liability Not a Liability? Textual Analysis, Dictionaries, and 10-Ks,' The Journal of Finance, vol. 66, no. 1, pp. 35-65, 2011.

[5] D. F. Larcker and A. A. Zakolyukina, 'Detecting Deceptive Discussions in Conference Calls,' Journal of Accounting Research, vol. 50, no. 2, pp. 495-540, 2012.

[6] L. Purda and D. Skillicorn, 'Accounting Variables, Deception, and a Bag of Words: Assessing the Tools of Fraud Detection,' Contemporary Accounting Research, vol. 32, no. 3, pp. 1193-1223, 2015.

[7] L. Loukas, M. Fergadiotis, I. Androutsopoulos, and P. Malakasiotis, 'EDGAR-CORPUS: Billions of Tokens Make The World Go Round,' Proc. 3rd Workshop on Economics and Natural Language Processing, pp. 13-18, 2021.

[8] J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, 'BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,' Proc. NAACL-HLT, pp. 4171-4186, 2019.

[9] D. Araci, 'FinBERT: Financial Sentiment Analysis with Pre-trained Language Models,' arXiv:1908.10063, 2019.

[10] Y. Yang, M. C. S. Uy, and A. Huang, 'FinBERT: A Pretrained Language Model for Financial Communications,' arXiv:2006.08097, 2020.

[11] W. Ping, Y. Jiao, H. Fan, and X. Zhang, 'Multimodal Fraud Detection in Financial Statements: A Trimodal Attention Network With Contrastive Evidence Chain Construction,' IEEE Access, vol. 14, pp. 80456-80468, 2026.

[12] T. Chen and C. Guestrin, 'XGBoost: A Scalable Tree Boosting System,' Proc. 22nd ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining, pp. 785-794, 2016.

[13] W. L. Hamilton, Z. Ying, and J. Leskovec, 'Inductive Representation Learning on Large Graphs,' Advances in Neural Information Processing Systems, vol. 30, pp. 1024-1034, 2017.

[14] P. Velickovic, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y. Bengio, 'Graph Attention Networks,' Proc. International Conference on Learning Representations, 2018.

[15] M. Schlichtkrull, T. N. Kipf, P. Bloem, R. van den Berg, I. Titov, and M. Welling, 'Modeling Relational Data with Graph Convolutional Networks,' The Semantic Web (ESWC 2018), pp. 593-607, 2018.

[16] Z. Hu, Y. Dong, K. Wang, and Y. Sun, 'Heterogeneous Graph Transformer,' Proc. The Web Conference 2020, pp. 2704-2710, 2020.

[17] Z. Liu, Y. Dou, P. S. Yu, Y. Deng, and H. Peng, 'Alleviating the Inconsistency Problem of Applying Graph Neural Network to Fraud Detection,' Proc. 43rd Int. ACM SIGIR Conf., pp. 1569-1572, 2020.

[18] Y. Dou, Z. Liu, L. Sun, Y. Deng, H. Peng, and P. S. Yu, 'Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters,' Proc. 29th ACM Int. Conf. Information and Knowledge Management, pp. 315-324, 2020.

[19] Yue, X., & Zhu, R. (2024). 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, 2(1), 63-72.

[20] Jiao, Y., & Liang, Y. (2024). An integrated process mining and explainable machine learning framework for bottleneck and internal control risk detection in the multi entity financial close. Social Science and Management, 1(3), 96-107.

[21] Bian, M., Wang, Y., & Lin, Y. (2024). Explainable supplier risk prediction for critical telecommunications infrastructure procurement: An empirical XGBoost–SHAP framework. Social Science and Management, 1(1), 64-77.

[22] Liang, Y., & Jiao, Y. (2024). Real-time financial fraud detection using a feature-light temporal dynamic graph neural network with contrastive learning. Journal of Trends in Finance and Economics, 1(1), 76–84.

[23] Shang, W., & Chen, T. (2024). PrefixKD-Edge: Real-time locomotion-intent prediction with prefix self-distillation on lightweight temporal networks. Innovation and Technology Studies, 1(1), 30–38.

[24] Rhee, M., Zou, J., & Qin, Y. (2024). CLARITY: A closed loop framework for duplicate aware failure prediction and budgeted automated remediation in cloud infrastructure. AI and Data Science Journal, 1(1), 76-85.

[25] Zhang, C., & Wang, Y. (2024). A data-driven multi-objective production scheduling framework for battery pack manufacturing under material arrival uncertainty. AI and Data Science Journal, 1(1), 86–93.

[26] Lin, Y., Wang, Y., & Bian, M. (2024). XOR EW: An explainable and cost aware artificial intelligence framework for operational risk early warning in financial institutions. AI and Data Science Journal, 1(1), 94-102.

Downloads

Published

2026-09-10