MCC-Vis: Visual Analysis for City Regional Co-occurrence Pattern Based on Traffic Trajectory Data

Co-occurrence between urban regions is a kind of mobile mode that focuses on inter-regional flow relations. Co-occurrence refers to the phenomenon that people from multiple regions visit the same urban area in a specific period, which is of high research value for business strategy and urban planning. Existing research on inter-regional co-occurrence mainly focuses on pattern recognition and event extraction, but there is few research on establishing inter-regional association through co-occurrence. Extracting co-occurrence can help users understand the relationship between human movement patterns in spatial and temporal views. This paper proposes a vectorized region representation method based on region co-occurrence by using graph embedding technology. A novel visual analysis system is designed based on vectorized representation, enabling users to analyze the relationship between urban areas. Finally, results of case study shows that the analyst can effectively analyze information implied by the co-occurrence phenomenon in urban areas and explore similar areas between cities by interacting with the system.