Taxi travel patterns and spatiotemporal heterogeneity: a POI data-based analysis

To improve urban taxi resource allocation and operational efficiency, this paper investigates the spatiotemporal characteristics and heterogeneous influences of taxi travel demand in Haikou using Points of Interest (POI) data and taxi order data. To analyze the impact of urban functional elements on taxi demand, we apply four regression models: Ordinary Least Squares (OLS), Geographically Weighted Regression (GWR), Multiscale Geographically Weighted Regression (MGWR), and Geographically and Temporally Weighted Regression (GTWR). Among these, MGWR shows superior performance, with weekday origin R² values improving by 13.1%, 7.7%, and 11.0% over OLS, GWR, and GTWR, respectively. The analysis identifies transportation facilities financial institutions and lifestyle services are key drivers of demand. While transportation facilities consistently exhibit positive correlations with taxi demand, the effects of financial institutions vary by location. These findings contribute to more precise taxi operation strategies and offer a data-driven basis for urban spatial planning.