This study presents a novel multi-modal approach analyzing public sentiment towards interactive art across Jiangsu Province using social media data. By integrating computer vision and NLP techniques (fine-tuned Qwen-VL for image captioning and prompt-based LLMs for sentiment analysis), we capture nuanced digital representations of artworks and public reactions. Findings reveal complex spatial patterns challenging traditional urban-rural dichotomies and highlighting Jiangsu’s polycentric cultural innovation. Urban centers focus on technological aspects and critical discourse, while peripheral areas emphasize thematic content and audience engagement. Correlation analysis reveals relationships between socioeconomic factors and digital art engagement, reflecting cultural capital and digital divide theories. These patterns invite re-examination of China’s cultural development through postreform urbanization theories, suggesting place-based policies that recognize diverse strengths across the urban-rural continuum. This research contributes to cultural democratization debates and offers a replicable framework for data-driven cultural policy-making.
Zhang et al. (Thu,) studied this question.