Semantic-Fused Multi-Granularity Cross-City Traffic Prediction: Enhancing Urban Mobility through Knowledge Transfer

The core message of this paper is to propose a Semantic-Fused Multi-Granularity Transfer Learning (SFMGTL) model that can effectively leverage knowledge from data-rich source cities to improve traffic demand prediction in data-scarce target cities. The model dynamically fuses multiple urban semantics, learns hierarchical node clustering, and extracts domain-invariant meta-knowledge to enable robust cross-city knowledge transfer.