A Causal–Priority Framework for Online Private Tour Satisfaction: Integrating LDA with DEMATEL–DANP

With the rapid growth of platform-based tourism, understanding how user perceptions influence satisfaction and decision-making priorities has become increasingly important for sustainable tourism governance. This study introduces a data-driven decision-support framework for assessing satisfaction with online private tours by combining Latent Dirichlet Allocation (LDA) with DEMATEL and DEMATEL-based ANP (DANP). Using user reviews, LDA extracts key satisfaction-related themes, while DEMATEL uncovers their causal relationships, and DANP ranks their importance within interdependent conditions. The findings indicate that satisfaction is primarily experience-driven, supported by upstream functional and contextual factors, with platform services serving as facilitators rather than direct drivers. By separating causal influence from priority ranking, this framework aids evidence-based resource allocation and governance-oriented decision-making in digital tourism systems. This study offers a transferable methodological approach for sustainability-focused tourism management and platform governance.