The impact of human mobility data scales and processing on movement predictability.

Area of Science:

• Computational Social Science
• Mobility Data Analysis
• Geospatial Intelligence

Background:

• Predictability of human movement is a theoretical upper bound for movement prediction models.
• Previous studies show variability in human mobility dataset predictability due to differing data processing.
• Limited research has explored the extent of this variability and its causes.

Purpose of the Study:

• To analyze how data representation impacts human movement predictability using high-precision trajectories.
• To investigate the influence of spatio-temporal scales and processing methods on mobility data properties.
• To understand the nonlinear dependencies between data representation and movement predictability.

Main Methods:

• Utilized high-precision individual movement trajectories.
• Applied various data processing and representation methods common in human mobility research over the last 11 years.
• Analyzed data across a wide range of spatio-temporal scales.

Main Results:

• Spatio-temporal resolution significantly impacts human movement predictability.
• Data processing methods demonstrably alter the predictability of mobility data.
• Nonlinear relationships were identified between data representation choices and the resulting data properties.

Conclusions:

• Data representation choices critically influence the predictability of human movement.
• Spatio-temporal resolution and processing methods are key factors affecting mobility data analysis outcomes.
• Understanding these impacts is crucial for accurate human mobility modeling and prediction.