Exploring Periods of Low Predictability in Daily Life Mobility

Researchers studying daily life mobility patterns have recently shown that humans are typically highly predictable in their movements. However, no existing work has examined the boundaries of this predictability, where human behaviour transitions temporarily from routine patterns to highly unpredictable states. Yet, this is arguably one of the most interesting and critical states, where users might be most in need of context-aware mobile applications. To address this shortcoming, we suggest a novel real-time estimator that can calculate an individual’s instantaneous entropy : a measure for their momentary predictability. Applying this to a rich dataset, we show that individuals display high variance in their predictability over time. Furthermore, we demonstrate that mobile application usage patterns are correlated with instantaneous entropy, thus indicating that people use applications differently in unfamiliar situations.