Understanding Human Mobility Characteristics Through Behavior and Corresponding Environmental Information

Understanding human mobility characteristics is essential for enhancing the functionality of map applications and navigation systems because the users have many mobility styles (e.g., driving straight down wide roads, walking slowly along nature-rich roads, and so on.). In recent years, trajectories have been employed to analyze human mobility. These studies have shown that two aspects are important in understanding human mobility characteristics: “behavioral characteristics (speed and direction of travel)” and “environmental characteristics of travel routes (information on roads and surrounding facilities)”. However, previous research has not addressed the explicit representation of individual user mobility-related characteristics. Therefore, this study aims to represent human mobility characteristics on a computer based on trajectories and related GIS data. It will be possible to handle the characteristics of human mobility on ICT services, leading to more personalized navigation. As a first step, this paper uses the movement trajectories of taxis in Porto and the corresponding GIS data to identify the features that represent the characteristics of each taxi driver. Then, using the representation of movement characteristics utilized in this study, we perform taxi driver identification and show that it significantly outperforms the accuracy of an existing method that uses only trajectories. In addition, this paper analyzes the contribution of features to identification and the features that are important for representing their mobility.