Cultural Tourism Scene Construction and Satisfaction Modeling in Traditional Villages: Integrating Multi-Stakeholder Theory and Machine Learning

Abstract

Based on a long-term follow-up study of traditional villages in the Wanjiang River Basin, this study conducted in-depth interviews with key stakeholders - tourists, community residents, tourism enterprises and local governments. Using the means-end chain theory (MEC), the soft ladder method and grounded theory coding method were used to explore the interactions and connections between multiple stakeholders under the drive of tourism interests and symbiotic collaboration. Specifically, the study analyzed how these stakeholders participated in the construction of cultural tourism scenarios in terms of local expressiveness, symbolic perception and comprehensive productivity. In order to further quantify and predict tourists' satisfaction in the comprehensive cultural tourism scenario, this study used machine learning models including decision trees, random forests, K nearest neighbors (KNN) and gradient boosting. Among them, the KNN model performed best in terms of accuracy and F1 value, highlighting its advantages in capturing subtle patterns of satisfaction-related features. In summary, this study constructed a multi-stakeholder cultural tourism scenario construction model that supports the sustainable and circular development of traditional villages.