Exploring the Impact of Environmental and Behavioral Factors on Emotional Experience in Themed Shopping Malls Using Social Media Data and Interpretable Machine Learning

In the context of the growing experience economy, themed transformation has emerged as a key strategy for shopping malls seeking to enhance offline consumer experiences. However, existing studies on environmental experiences have largely focused on static factors, overlooking the potential interaction effects between the environment and experience behaviors. Based on 23,541 comments posted in 2024–2025 about 16 highly recommended Chinese-themed shopping malls on social media platforms, this study used natural language processing (NLP) to construct lexicons covering 11 environmental categories and 7 behavioral categories. The Light Gradient Boosting Machine (LightGBM) model, combined with SHapley Additive exPlanations (SHAP), was then employed to assess the relative importance of the environmental and behavioral elements and to reveal their interaction effects on customer sentiment. The results show that, among environmental elements, decoration is the most important predictor of sentiment, and among behavioral elements, leisure contributes the most. Leisure performs the strongest interactions with decoration and store category, suggesting that these couplings are most closely associated with positive emotional experience. These findings point toward a design logic centered on environment–behavior interdependencies, help to propose differentiated optimization strategies for various themed shopping malls, and offer a new perspective on the complex interplay among environment, behavior, and emotional experience.