Quantifying informal street vendors clusters through geo-spatial indicators and geo-tagged social media: a socio-spatial fusion approach for human-centered urban renewal

Amid rapid urbanization, informal vendors have become vital to the urban informal economy, with their spatial distribution shaped by market demand, social vitality, and governance boundaries. Yet existing studies often neglect dynamic social data and spatial heterogeneity, limiting support for fine-grained inclusive governance. Grounded in Lefebvre’s “spatial triad,” this study constructs a “market – society – governance” framework and examines informal vendors’ spatial mechanisms by integrating built environment indicators with social media check-in data. Using Suzhou, China, as a case, we employ OLS, MGWR, and XGBoost to analyze multi-source data and identify factor intensity and spatial variability. Results show: (1) vendors cluster in high-density, high-consumption, high-FAR, and transit-accessible areas, with OLS highlighting social perception via online data; (2) MGWR reveals strong spatial heterogeneity in effects of housing price, enforcement proximity, and bus distance, reflecting uneven governance and cost constraints; (3) XGBoost underscores nonlinear threshold effects of population density, consumption, building density, and entertainment facilities, indicating trade-offs between regulatory pressure and market opportunity. The study transforms Lefebvre’s spatial triad into a quantifiable governance framework by combining geotagged social-media data that capture perceived space with a hybrid MGWR – XGBoost approach, enabling the joint identification of spatial heterogeneity and non-linear regulatory thresholds.