Predictive and Resilience Construction in Tourism Risk and Crisis Management Driven by Artificial Intelligence

Abstract

The tourism industry is highly vulnerable to a wide range of risks and crises, including natural disasters, public health emergencies, geopolitical instability, and operational disruptions. The increasing frequency and complexity of such events have exposed limitations in traditional tourism risk and crisis management approaches, which are often reactive, fragmented, and experience-driven. Against this backdrop, artificial intelligence (AI) offers new possibilities for enhancing risk prediction, early warning, and resilience building in tourism systems.This study develops a theory-driven and mechanism-based framework to examine how three core AI approaches—natural language processing (NLP), knowledge graphs, and time-series models—can be integrated into tourism risk and crisis management. The analysis focuses on three innovation domains: risk prediction and early warning, resilience-oriented tourism supply chain construction, and the intelligent upgrading of emergency management systems. The study argues that AI contributes to tourism resilience not merely through automation, but by restructuring information processing, inter-organizational coordination, and decision-making under uncertainty.By linking AI techniques with tourism risk management theory and resilience thinking, this paper advances the literature on smart tourism and crisis management and provides practical insights for building more adaptive and resilient tourism systems in an era of increasing uncertainty.