UrbanMotion: Visual Analysis of Metropolitan-Scale Sparse Trajectories

Area of Science:

• Urban computing
• Geographic Information Systems (GIS)
• Data Visualization

Background:

• Visualizing large-scale human movement is crucial for urban problem-solving, including traffic optimization and business site selection.
• Existing trajectory visualization techniques struggle with temporally sparse mobile location data, characterized by long intervals between records.

Purpose of the Study:

• To introduce UrbanMotion, a visual analytics system designed to address the challenges of visualizing sparse, long-tailed trajectory data from mobile apps.
• To extend existing wind map designs with features for map-matched local movements, multi-directional population flows, and population distributions.

Main Methods:

• Developing effective methods to extract and aggregate population movements from dense trajectory segments, leveraging long-tailed sparsity.
• Extending the wind map design to incorporate map-matched local movements and multi-directional population flows.
• Utilizing a large-scale mobile location dataset covering millions of city residents.

Main Results:

• UrbanMotion successfully visualizes characteristic and anomalous patterns in urban population movement.
• The system supports analysis of commuting patterns, event detection, and business site configuration.
• Case studies and expert feedback confirm the system's significance and effectiveness.

Conclusions:

• UrbanMotion provides a powerful tool for analyzing and understanding urban human mobility from sparse mobile data.
• The system's novel approach enhances the capabilities of visual analytics for urban planning and management.
• Effective visualization of population movement is key to addressing complex urban challenges.