An approach to exploring the spatial distribution and influencing factors of urban problems based on Land use types

Numerous urban problems seriously affect the residents' well-being and urban governance. Urban problems are strongly correlated with land use types and are spatially uncertain. Existing research on urban problems and land use types is limited, predominantly relying on land planning data and focusing on specific problems. Therefore, using Sanya as an example, this study identified the latest land use types by graph convolutional network (GCN) model and major urban problems from 480170 "12345" hotlines, further explored spatial heterogeneity and influencing factors of multi-category urban problems at the street scale under land use types. The results showed social security, noise, and household registration were the three most severe problems. Urban problems' spatial distribution differed significantly among land use types. Illegal parking and household registration were predominantly concentrated in residential and commercial lands, with noise pollution in green lands, land requisition demolition and wage arrears in agricultural lands, and social security and education problems in public service lands. Land use types and points of interest (POIs) significantly impacted urban problems. Residential and commercial lands were the primary epicenter, influencing various urban problems. Green lands greatly impacted noise pollution. Healthcare POIs positively impacted illegal parking, while financial institution POIs negatively influenced social governance.