Transforming tourism sustainability through Artificial Intelligence: a multidimensional framework for overtourism prediction and mitigation

This research investigates the application of artificial intelligence for sustainable tourism management, with particular focus on the overtourism phenomenon and the development of predictive systems for risk monitoring in tourist destinations. The work proposes an innovative approach that integrates big data analytics, machine learning algorithms, and multidimensional assessment methodologies to create decision support tools for sustainable tourism planning. Tourism has emerged as one of the world's most significant economic forces, contributing approximately 10% to global GDP and serving as a primary driver of employment and infrastructure development across numerous destinations. However, this unprecedented growth has introduced critical sustainability challenges, most notably overtourism, a phenomenon where visitor numbers exceed a destination's sustainable capacity, resulting in environmental degradation, cultural dilution, infrastructure strain, and declining quality of life for local residents. Traditional tourism management approaches, including visitor caps, tourist taxes, and promotional diversions to alternative destinations, have proven insufficient in addressing the multidimensional complexity of this challenge. Simultaneously, the tourism industry contributes approximately 8% of total greenhouse gas emissions, with domestic air travel accounting for nearly half of this ecological footprint. This dissertation investigates how advanced analytical frameworks, particularly those leveraging artificial intelligence (AI), can provide transformative solutions to both overtourism and tourism's broader environmental impact. From an academic and disciplinary perspective, this study is positioned at the intersection of tourism management, sustainable development, and computational intelligence—areas that require a renewed methodological approach to address the multidimensional complexity of contemporary tourism systems. Traditional tourism management methodologies, based on visitor caps, tourist taxes, and spatial diversification, have proven inadequate for managing the dynamic and interconnected nature of modern tourism flows. The research contributes to the theoretical corpus of sustainable tourism management by introducing advanced analytical frameworks that overcome the limitations of static assessments, opening new perspectives for predictive understanding of tourism phenomena and optimization of destination management strategies. The primary objectives of this research are: (1) to develop an innovative methodological framework that integrates artificial intelligence and predictive analytics for overtourism risk assessment; (2) to design and validate the Overtourism Risk Index (ORI), a multidimensional index incorporating demographic, spatial, ecological, and infrastructural variables; (3) to demonstrate the effectiveness of real-time monitoring systems and personalized recommendation algorithms in reducing tourism's 13 ecological footprint. The work aims to bridge the existing methodological gap between theoretical research in sustainable tourism and practical implementation of technologically advanced solutions, providing sector stakeholders with operational tools based on empirical evidence for proactive and data-driven destination management. The expected scientific advancement consists in developing a methodological paradigm that transforms tourism management from reactive to predictive, enabling early identification of risk situations and implementation of personalized preventive measures. The research results will provide: a replicable model for quantitative assessment of overtourism risk; machine learning algorithms optimized for tourism flow prediction; evidence-based guidelines for integrating AI technologies into destination management processes; and methodologies for evaluating the impact of implemented mitigation strategies. This approach will contribute significantly to the transition towards more sustainable tourism, balancing economic growth, environmental preservation, and local community well-being. The structure of this dissertation reflects a progressive approach that leads from theory to practice through four interconnected chapters that progressively build from theoretical foundations to practical applications. Chapter 1 presents the theoretical foundation and evaluates existing literature review examining analytical models in tourism management, with particular emphasis on overtourism. This chapter traces the evolution from traditional economic and GIS-based approaches to modern AI- driven methodologies, identifying critical limitations in conventional models and highlighting how machine learning, natural language processing, and predictive analytics represent paradigm shifts in tourism management capabilities. Chapter 2 explores AI's transformative potential in reducing tourism's ecological footprint through detailed case studies of Rome and Venice, demonstrating how AI interventions throughout the tourist journey—from pre-trip planning to post-visit feedback—can enhance sustainable travel decisions, reduce ecological footprints, and enable proactive, evidence- based destination management that balances economic benefits with environmental preservation and community well-being. Chapter 3 introduces the Overtourism Risk Index (ORI), a comprehensive multidimensional framework integrating demographic pressure, spatial congestion, ecological vulnerability, and infrastructure strain, enhanced through AI-driven predictive analytics, real-time monitoring systems, and personalized recommendation engines that offer unprecedented opportunities to optimize visitor flows, reduce ecological footprints, and enable evidence-based destination management. Chapter 4 provides a critical assessment of current gaps in tourism management methodologies and demonstrates how the proposed AI-driven approaches address these limitations while establishing foundations for future research directions.