CausalPOI: Spatio-Temporal Graph-Based Causal Modeling for Cold-Start POI Check-in Forecasting
Researchers have developed CausalPOI, a new framework for forecasting check-in patterns at new Points of Interest (POIs) by modeling their spatio-temporal dynamics and causal relationships with existing POIs. This approach utilizes a Spatio-Temporal Functional Interaction Graph to capture semantic and spatial connections, and simulates factual and counterfactual scenarios to estimate causal effects. Experiments on real-world data show CausalPOI significantly outperforms existing methods in forecasting and urban intervention analysis. AI
IMPACT Provides a more interpretable and actionable foundation for urban planning and commercial decision-making by improving POI forecasting.