The Impact of an Artificial Intelligence-Based Forecasting Model on the Development of Sustainable Tourism

This research investigates the application of artificial intelligence (AI) and Web 3.0 technologies in promoting sustainable urban tourism, with a particular focus on demand forecasting and environmental impact assessment. The study presents a two-layered AI-based model aimed at supporting data-driven decision-making in destination management, addressing the need for forward-looking strategies that align with both operational and sustainability goals. The research applies the Facebook Prophet algorithm to forecast monthly tourism demand in two Hungarian cities—Budapest and Győr—selected for their contrasting tourism profiles. Forecast outputs were then integrated into a sustainability impact module estimating carbon dioxide emissions, water consumption, and waste generation, based on empirically defined conversion factors. Results indicated strong seasonal peaks in Budapest, with over 1.3 million overnight stays projected for August 2026, and corresponding environmental impacts surpassing 62,000 tons of CO2. In contrast, Győr exhibited more moderate fluctuations and lower error margins, reflecting a more stable tourism pattern. Forecast accuracy was assessed using MAE, RMSE, and MAPE metrics, showing acceptable performance for strategic use, although with reduced reliability in low-demand periods. The sustainability module effectively highlighted peak periods of ecological burden, enabling targeted interventions such as infrastructure scaling, service optimization, and seasonal policy adjustments. In addition to its forecasting functionality, the model offers practical guidance for municipalities by identifying where and when ecological pressure is likely to arise. The dual-model framework offers a scalable and replicable approach for cities seeking to balance tourism growth with environment al and community well-being. By integrating predictive analytics with sustainability assessment, the model provides valuable insights into the timing and magnitude of tourism’s impact. This supports smarter capacity planning, emission reduction strategies, and the alignment of visitor flows with local resilience thresholds. The findings contribute to the evolving discourse on smart and sustainable tourism in the Web 3.0 era, positioning AI as a critical enabler of holistic and proactive destination manageme nt