Publication

Abstract : Collaborative learning is widely accepted as an approach to promote learning effectiveness and student satisfaction. However, the quality and outcomes of collaboration depend upon a number of factors, among which group formation plays an important role. Existing approaches take into account groups formed through random assignment or based on certain criteria such as academic performance, demographic features, and communication skills. This paper proposes an approach where preferences of learners in terms of the composition of the group to which they would like to belong are taken into account. Machine learning algorithm (K-Means) is applied to cluster large learner groups and to analyze their preferences. Further, prior knowledge and communication skills of learners are likely to influence effectiveness of collaborative learning. Therefore, these two factors are considered in addition to learner preferences. Initially, groups are formed using learner preferences. Then an attempt is made to include learners with good academic performance and communication skills in each group. Results from this study shows that the proposed approach for group formation enhances learner satisfaction in collaborative learning, which is likely to improve learning outcomes.