Ensuring positive emotional experiences for tourists is crucial for rural development sustainability, yet research often neglects the rural built environment. This study fills this gap by investigating the impact of the rural built environment on tourist emotions, focusing on traditional villages in Fuzhou, China. Natural Language Processing (NLP) techniques analyze web text data to determine sentiment indices. The study utilizes HRnet segmentation and Matlab for built environment data collection, and employs eXtreme Gradient Boosting (XGBoost), the SHapley Additive exPlanation (SHAP) model, and ArcMap software for analysis. Findings reveal significant spatial variation in sentiment indices, with Houfu, Qianhu, and Ximen villages scoring highest, and Doukui, Jiji, and Qiaodong villages scoring lowest. Key environmental indicators impacting sentiment include Openness, Greenness, and Color Complexity, with non-linear relationships observed. The study offers insights for sustainable traditional village development, informing effective design strategies to enhance tourist sentiment in rural settings.
