Enhancing cultural tourism performance through data-driven evaluation: a hybrid MCDM–machine learning approach in Northern Thailand

Abstract

Heritage-rich regions increasingly rely on cultural tourism as a driver of preservation, economic growth, and inclusive development. This study presents an integrated, data-driven framework to evaluate and enhance the performance of cultural tourism destinations. By combining Analytic Hierarchy Process (AHP)-weighted Multi-Criteria Decision-Making (MCDM) techniques–Simple Additive Weighting (SAW), Weighted Product Model (WPM), and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS)–with K-Means clustering and Decision Tree analysis, the study assesses 125 destinations based on six key criteria: Satisfaction, Flexibility, Sustainability, Safety, Economic Impact, and Cultural Conservation. Sites were grouped into Low, Medium, and High performance clusters, and interpretable decision rules were derived to guide strategic improvements. For example, safety scores above 6.73 and cultural conservation above 6.06 were key predictors of high performance. The model achieved 90.4% classification accuracy and 80% cross-validated accuracy. This hybrid framework not only provides a robust evaluation of destination performance but also generates actionable insights for tourism planners and policymakers. It supports sustainable tourism development by offering clear guidelines for enhancing underperforming sites, thereby contributing to the effective integration of data science in tourism and recreation planning.