DIGITAL TRANSFORMATION OF TAXATION PROCESSES BASED ON A DATA-DRIVEN APPROACH: CONCEPT AND ARCHITECTURE OF AN INTEGRATED ANALYTICAL CONTOUR

Abstract

The purpose of the article is to substantiate the theoretical and methodological foundations of the digital transformation of taxation processes based on a data-driven approach and to develop a conceptual model of an integrated analytical contour for tax liability management. The research is based on a combination of general scientific and specialized methods, including system analysis, comparative analysis, classification, and methodological abstraction. These methods were used to identify the structural components of digital tax administration, analyze traditional and data-driven approaches, and synthesize a unified analytical framework. The study demonstrates that digital transformation fundamentally changes the logic of tax administration, shifting from retrospective control to proactive, analytics-driven management. It identifies Big Data infrastructure as the core operational component enabling continuous data processing and integration. The effectiveness of machine learning models, particularly ensemble methods such as Random Forest and XGBoost, is confirmed in tax risk prediction, achieving accuracy levels up to 92–93%. The integration of XAI approaches ensures interpretability and regulatory compliance of algorithmic decisions. The scientific novelty lies in the development of an integrated digital analytical contour model (IDAC-TL), which combines five functional levels and five methodological principles, including continuity, full coverage, adaptability, explainability, and institutional embeddedness. This model provides a comprehensive framework for integrating accounting, analytical, and management subsystems within a unified digital environment. The proposed approach enables optimization of tax administration processes, reduction of compliance costs, and improvement of decision-making quality. It also supports the transition to risk-oriented and automated control mechanisms, enhancing the efficiency of resource allocation and reducing corruption risks. The study concludes that the implementation of a data-driven approach in taxation requires not only technological advancements but also institutional and regulatory transformation. The proposed model can serve as a methodological basis for further research and practical implementation in different economic sectors, particularly in the context of digital transformation in Ukraine.