An initial study on college students' daily activities using GPS trajectories

As GPS trajectories trace the moving objects in space and represent the mobility of individual carriers in real-world, it provides a new method for researchers to study human activities in urban time-space geography. This paper proposes a referential workflow to explore the pattern of college students' activities based on GPS trajectories and gives a trial experiment. We divide the trajectories into stay-and-move sub-trajectories, and analyze the potential trend of activities based on the attributes of the stay points as well as some basic information of the carriers. After comparing with the results collected from the traditional questionnaire, we approach the conclusion that GPS trajectories could be very beneficial for human behavior mining.