A Process Mining approach to analyze learning behavior in the flipped classroom

Abstract

Flipped classroom as an instructional strategy promotes dynamic learning to increase student's engagement and learning outcomes. Video-based lectures and materials are to be referred before the actual class as a pre-class activity in the flipped classroom setting and completion status are evaluated through learner's feedback. In this paper, pre-class activities are evaluated by applying process mining on the student's clickstream data generated through learning platform “Delta” designed for current work. The students are categorized as credible and mediocre performers based on their pre-assessment score and process models are generated by applying fuzzy miner and heuristic miner based on their engagement with the tool. We further analyzed and registered the absolute & relative occurrences of events, identified a contrasting pattern in student behavior through trend analysis and process comparator. The various process models from learner's data were presented using several visualizations.