Modern tax administration is undergoing a transition from formal inspection procedures to intelligent analytical models based on digital data and artificial intelligence technologies. The purpose of the study is to develop a methodology for constructing an integrated tax risk index that ensures accurate, dynamic, and interpretable assessment of taxpayer behavior. The proposed three-level model combines behavioral, network, and semantic analyses, allowing for the integration of both quantitative and contextual indicators of risk. The index is calibrated on real outcomes and employs explainable AI techniques to interpret the contribution of each factor. Scientific novelty lies in integrating heterogeneous sources of information into a unified analytical framework adaptable to changes in legislation and data structures. Practical significance is determined by the applicability of the proposed model in tax monitoring and corporate compliance systems for early detection of deviations and enhancing the accuracy of decision-making
tax risk; integrated index; risk-based approach; explainable artificial intelligence; behavioral analysis; counterparty network analysis; semantic analysis (legal context); tax monitoring
1. Berezina A.D., Grundel L.P. Analiz nalogovyh riskov pri formirovanii nalogovoy strategii organizacii // Russian Economic Bulletin. 2022. T. 5. № 3. S. 200–203.
2. Kuz'mina A.A., Variativnost' i rezul'tativnost' metodov ocenki regional'nogo nalogovogo potenciala // Mir ekonomiki i upravleniya. 2024. T. 24. № 1. S. 71–81.
3. Nalogovyy kontrol': pravovye problemy i riski / pod red. S. G. Pepelyaeva; sost. S. A. Sosnovskiy. – Moskva : Statut, 2025. – 400 s.
4. Pimenov N. A.; Demin, S. S. Nalogovoe planirovanie: uchebnik dlya vuzov. – 4–e izd., pererab. i dop. – Moskva: Izdatel'stvo Yurayt, 2025. – 137 s.
5. Ruban-Lazareva N.V. Nalogovye riski v kontekste dostizheniya nacional'nyh celey razvitiya Rossii na 2024 god // Ekonomika. Nalogi. Pravo. 2024. T.17. № 1. S. 145–157.
6. Tagil'ceva A.V., Semikolenova M.N. Upravlenie nalogovymi riskami v sisteme ekonomicheskoy bezopasnosti organizacii V knige: Sovremennye aspekty ucheta, analiza i audita. Materialy Vserossiyskoy nauchno-prakticheskoy konferencii. Krasnoyarsk, 2025. S. 80–84.
7. Chau G., Leung P. A critical review of Fischer tax compliance model: A research synthesis // Journal of Accounting and Taxation, 1 (2) (2009), pp. 034–040.
8. Lundberg S. M., Lee S.–I. A Unified Approach to Interpreting Model Predictions // Advances in Neural Information Processing Systems 30 (NIPS 2017). URL: https://arxiv.org/pdf/1705.07874 / (data obrascheniya: 17.10.2025).
9. OECD. Tax Administration 2023: Comparative Information on OECD and other Advanced and Emerging Economies. Paris: OECD, 2023. URL: https://www.oecd.org/en/publications/tax–administration–2023_900b6382–en.html / (data obrascheniya: 17.10.2025).
10. Ribeiro M. T., Singh S., Guestrin C. «Why Should I Trust You?»: Explaining the Predictions of Any Classifier. KDD 2016. URL: https://arxiv.org/pdf/1602.04938 / (data obrascheniya: 10.10.2025).
11. Zheng Qinghua Yiming Xu, Huixiang Liu, Bin Shi, Jiaxiang Wang, Bo Dong. A Survey of Tax Risk Detection Using Data Mining Techniques //Engineering, Volume 34, 2024, Pages 43–59
