Process Mining based Student's Behavioural modeling for Online Assessment

In recent times the technology has paved the path for online tests and evaluations in the contemporary educational environment. The educational institutes all over the world understand that instructing and assessing pupils are crucial. These tasks are not limited to the walls of a conventional physical classroom and can be carried out whenever it is convenient regardless of where the teacher or the student is located with the use of smart devices, laptops, and personal computers. One of the less explored paths in an online assessment was to understand the behavior of students during the assessment. This research focuses on understanding the behavior of students during online assessment using process mining technique. It was observed that trace clustering approach was the most efficient method in grouping students as per their interaction behavior with assessment tool with silhouette score of 0.68 for the cluster size of 4. The process model for various clusters formed through trace clustering was compared with the ideal model and the study also concluded the stochastic similarity between the groups with the ideal model.