Usage pattern recognition in student activities

This paper presents an approach of collecting contextualized attention metadata combined from inside as well as outside a LMS and analyzing them to create feedback about the student activities for the teaching staff. Two types of analyses were run on the collected data: first, key actions were extracted to identify usage patterns and tendencies throughout the whole course and then usage statistics and patterns were identified for some key actions in more detail. Results of both analyses were visualized and presented to the teaching staff for evaluation.