Sustainable data-driven framework and policy recommendations for enhancing sports promotion using generative and explainable Artificial Intelligence

Abstract

Problem Statement: Sports are fundamentally crucial for improving physical health and strengthening community bonds. However, the current policy framework faces several challenges, including considerable imbalances in sports promotion across regions and age groups, insufficient data, and the absence of scientific evidence to guide policy design. These issues hinder the realization of fair and sustainable sports policies. Approach: This study proposes a data-driven framework that integrates generative and explainable artificial intelligence. This framework leverages generative artificial intelligence models wrapped in the synthetic data vault library to address data scarcity and bias. By analyzing statistical data related to sports promotion, the proposed framework would facilitate comprehensive policy recommendations that account for regional characteristics and intergenerational differences. Purpose: This study aims to establish a scientific foundation to support two primary policy goals: enhancing competitive performance and promoting health, thereby improving the fairness and effectiveness of sports policies via this framework. Results: This study evaluated the potential of generative artificial-based data supplementation in designing exercise programs tailored to regional characteristics and generational diversity through hypothetical applications. The use of explainable artificial intelligence enhances the transparency of the analytical results, increasing the credibility of policy recommendations. This framework offers a starting point for the theoretical examination of the applicability of data-driven approaches as a foundation for policy design. Conclusion: The proposed framework addresses challenges in sports policy, with potential applicability in other fields, such as healthcare and education. In addition, by addressing ethical issues, including bias and privacy protection, enhancing the reliability of datadriven policy designs is essential. Future research is expected to validate the effectiveness of the framework through empirical studies, thereby contributing to the realization of sustainable policies based on scientific evidence.