Profiling mobility patterns and driving behaviors of individual drivers via trajectory trait

<p>Driver profiling can provide a human-centered approach to portraying individual travel behavior and revealing their motivation, objectives, and needs, thereby contributing to driving safety analysis, location-based service, and intelligent transportation. However, existing trajectory-based methods are limited to measuring low-level features, such as average speed and radius of gyration. Although these features can characterize specific observable behaviors, such as driving operation and movement range, they fail to depict stable traits underlying individual travel behavior. In this study, inspired by the Big Five Personality Traits, we model the driver profile through four fundamental trajectory traits: extroversion, openness, neuroticism, and conscientiousness, and quantify these traits by developing a Trajectory Trait Scale (TTS). Experiments on more than one million trajectories from 2,051 anonymized private vehicle volunteers over eight months demonstrate that our method can provide a valid representation of individual drivers’ mobility patterns and driving behaviors. Specifically, we validate the consistency between trajectory traits and vehicle customer service records of drivers, including life rescue, navigation service, violation query, and fatigue companion. Besides, we find that trajectory integrity, seasonal changes, and traffic conditions exert small but non-negligible impacts on the stability of trajectory traits. These findings can enhance the understanding of human behavior in various spatiotemporal contexts, and illuminate the relations between trajectory traits and personality traits.</p>