Analyzing the group formation process in intelligent tutoring systems

Abstract

Due to the increasing use of active learning methodologies such as the Flipped Classroom, which usually use group activities, Intelligent Tutoring Systems (ITSs) have the opportunity of supporting them. In this work, we present a group formation tool that allows teachers to create easily and efficiently groups of students based on the K-means algorithm. Our tool allows creating the groups using videos’ and exercises’ information. Moreover, we have validated the suitability of our group formation process and tool through the students’ scores. In order to gain insights into the group formation in learning systems which only use exercises or use mainly videos, we also compare the groups through the Jaccard similarity measurement. Results show that the groups formed using only the videos’ information are quite similar to the groups created using just the exercises’ information. Therefore systems that only use exercise interactions for group formation might be replaced by others that only use video interactions and the other way around.