Exploring the association between mobility behaviours and academic performances of students: a context-aware traj-graph (CTG) analysis

Analysing the mobility traces of moving agents (mobile users, GPS-equipped vehicles, CDRs, etc.) may help in interpreting the “human interests and intentions” behind the movements and thus facilitates diverse range of location-based applications. The trajectory analysis uncovers the connections, correlations and differences among individuals and their activities by exploring their mobility attributes. This paper focuses on how mobility information (GPS traces) of a student exhibits correlation with her academic performance. The proposed framework analyses the GPS trajectories of students in an academic campus, models the mobility patterns of students using context-aware traj-graph (CTG), clusters signature mobility patterns and uncovers the correlation of mobility attributes with the academic performance of the students. A mobility knowledge graph has been constructed considering the entities, namely students, places of visits, movement behaviours, subjects, academic performances and the relationships among the entities. Using real-life dataset of an academic campus, we demonstrate that the mobility attributes are associated with students’ academic performances and students’ academic performance can be predicted from their movement behaviours.