Abstract
This study aimed to evaluate the performance of artificial intelligence strategies in detecting accounting anomalies and their contribution to the efficiency of review processes. A quantitative experimental study was conducted using synthetic accounting records from two fiscal years. Four strategies were compared: traditional rules, Isolation Forest, Random Forest, and a hybrid model. Performance was assessed using average precision, area under the ROC curve, and anomaly coverage under different inspection budgets. Random Forest achieved the best overall performance, with an average precision of 0.5392 and a ROC-AUC of 0.9517. Under a 2% review budget, it identified 379 out of 427 anomalies (88.76%). The results show the potential of artificial intelligence to prioritize accounting records, however, its organizational application is suggested after having been empirically verified.
References
[2] N. T. T. Thuy, “Effect of accounting information system quality on decision-making success and non-financial performance: Does non-financial information quality matter?,” Cogent Business & Management, vol. 12, no. 1, Art. no. 2447913, 2025, doi: 10.1080/23311975.2024.2447913.
[3] J. Kokina and T. H. Davenport, “The emergence of artificial intelligence: How automation is changing auditing,” Journal of Emerging Technologies in Accounting, vol. 14, no. 1, pp. 115–122, 2017, doi: 10.2308/jeta-51730.
[4] D. A. Appelbaum, A. Kogan, and M. A. Vasarhelyi, “Analytical procedures in external auditing: A comprehensive literature survey and framework for external audit analytics,” Journal of Accounting Literature, vol. 40, pp. 83–101, 2018, doi: 10.1016/j.acclit.2018.01.001.
[5] J. Perols, “Financial statement fraud detection: An analysis of statistical and machine learning algorithms,” Auditing: A Journal of Practice & Theory, vol. 30, no. 2, pp. 19–50, 2011, doi: 10.2308/ajpt-50009.
[6] Y. Bao, B. Ke, B. Li, Y. J. Yu, and J. Zhang, “Detecting accounting fraud in publicly traded U.S. firms using a machine learning approach,” Journal of Accounting Research, vol. 58, no. 1, pp. 199–235, 2020, doi: 10.1111/1475-679X.12292.
[7] J. Davis and M. Goadrich, “The relationship between precision-recall and ROC curves,” in Proc. 23rd Int. Conf. Machine Learning, Pittsburgh, PA, USA, 2006, pp. 233–240, doi: 10.1145/1143844.1143874.
[8] T. Saito and M. Rehmsmeier, “The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets,” PLOS ONE, vol. 10, no. 3, Art. no. e0118432, 2015, doi: 10.1371/journal.pone.0118432.
[9] J. Gronewald, S. Stephan, A. Rombach, and P. Fettke, “Generating synthetic journal entries for audit analytics: Method and dataset,” Data in Brief, vol. 69, Art. no. 113257, 2026, doi: 10.1016/j.dib.2026.113257.
[10] L. Breiman, “Random forests,” Machine Learning, vol. 45, no. 1, pp. 5–32, 2001, doi: 10.1023/A:1010933404324.
[11] F. T. Liu, K. M. Ting, and Z.-H. Zhou, “Isolation forest,” in Proc. 8th IEEE Int. Conf. Data Mining, Pisa, Italy, 2008, pp. 413–422, doi: 10.1109/ICDM.2008.17.
[12] F. T. Liu, K. M. Ting, and Z.-H. Zhou, “Isolation-based anomaly detection,” ACM Transactions on Knowledge Discovery from Data, vol. 6, no. 1, Art. no. 3, 2012, doi: 10.1145/2133360.2133363.
[13] M. Schreyer, T. Sattarov, D. Borth, A. Dengel, and B. Reimer, “Detection of anomalies in large scale accounting data using deep autoencoder networks,” arXiv:1709.05254, 2017, doi: 10.48550/arXiv.1709.05254.
[14] J. Gronewald, S. Stephan, A. Rombach, and P. Fettke, “DIB General Ledger Data,” Zenodo, 2026, [Data set], doi: 10.5281/zenodo.22014653.

This work is licensed under a Creative Commons Attribution 4.0 International License.
