Assessing urban renewal efficiency via multi-source data and DID-based comparison between historical districts

This study evaluates the effectiveness and mechanisms of two historic district renewal (HDR) models in East China: Nanjing’s Xiaoxihu-Laomendong District (XLD) and Suzhou’s Shiquanjie Street District (SSD), by using multi-sourced socio-spatial data, machine learning-based text semantic segmentation, and the Difference-in-Differences (DID) model. XLD’s government-driven model integrates systematic relocation and stakeholder collaboration, while SSD’s market-driven model emphasizes rapid commercial activation with minimal displacement. Both renewal models enhance spatial vitality and public expressed sentiment. SSD, driven by economic activation and stakeholder collaboration, enables rapid revitalization in areas with high development potential. While XLD, with its resident-centered, gradual approach, fosters long-term community resilience and cultural preservation. The effectiveness of each model depends on the district’s socio-economic and cultural context. XLD’s increased facility-related POIs reflect enhanced spatial aggregation, while SSD’s focus on commercial POIs accelerates revitalization. Renewal effectiveness varies with socio-economic and cultural contexts, offering insights for context-specific urban renewal strategies and policy.