Dynamic construction and maintenance of destination competitiveness based on interpretable machine learning

Abstract

This study proposes a theoretical framework for dynamic destination competitiveness (DD-TDC) by integrating digital empowerment and dynamic governance mechanisms. Using data on the competitiveness of 119 destinations from the World Economic Forum (2018–2023), the Interpretable Random Forest Model and SHapley Additive exPlanations analysis was conducted to explore the drivers of competitiveness during crisis and non-crisis periods. The findings reveal that (1) Policy factors exhibit a short-term dominant effect on competitiveness through dynamic governance mechanisms during the COVID-19 pandemic; (2) in the long term, infrastructure is the most significant driver of competitiveness, followed by tourism resources and the business environment, while policy and sustainability contribute relatively less; and (3) the long-term impact of COVID-19 on destination competitiveness is limited. This study validates the dynamic adaptability of the DD-TDC model and provides methodological insights for destination managers to address crises and build sustainable competitiveness. The results have valuable implications for both academic research and practical destination management strategies.