This study explores the social and perceptual impacts of urban renewal in Ningbo, China, comparing three redevelopment types: urban village redevelopment, future community construction, and old residential community renovation. Using Douyin short videos (predominantly user-generated, with a minority of official/news/promotional clips), spoken narratives were analyzed and visual cues to characterize platform-visible emotional responses and salient themes. This study applies automatic speech transcription (validated on a manually coded subset) and uses computer vision to extract object/scene labels, enabling a comparative interpretation across project types. Future community videos are largely optimistic, emphasizing modern amenities and aspirational lifestyles. Old community renovation elicits mixed reactions: residents appreciate improvements yet frequently express frustration over construction-related inconvenience and everyday frictions. Urban village redevelopment is often supported in principle but is accompanied by concerns about relocation fairness and the preservation of community identity. By combining textual sentiment with visual context, this exploratory multi-modal analysis highlights nuanced emotional landscapes that conventional surveys may overlook. The approach may complement established participation and feedback channels by helping policymakers identify issue salience and tailor communication and engagement strategies across renewal types, while recognizing the representativeness constraints of platform data.
Liu et al. (Fri,) studied this question.