Mapping Multisensory Tourism in the Greater Bay Area: The Micro and Macro Perspectives

Understanding how tourists perceive destinations through their senses and how these processes shape affective responses is vital for experience-based tourism development. This study constructs and tests a resource – senses – sentiment model that bridges tourism resource types, multisensory impressions, and affective responses. Drawing on user-generated reviews from Chinese platforms, a dual-scale approach was employed to analysis: latent Dirichlet allocation was used to extract tourism resource themes, and a predefined five-sense vocabulary captured sensory dimensions. Sentiment was measured via BERT-based classifier. At the micro level, hypothesis testing validated the sequential effects of resources on sensory mentions and of sensory richness on sentiment types. At the macro level, sensory impact index was introduced to map city-level sensory strengths and compared patterns across resources and sentiments. Our study offers a scalable framework for linking environmental attributes with affective outcomes. The findings enrich theoretical comprehension of embodied tourism experiences and provide new tools for sensory – sentiment destination planning.