Ability based Collaborative Pairing and Grouping among Learners

It is well recognized that collaborative learning enhances academic performance by pooling the diverse strengths of co-participants and by providing a healthy competitive learning environment. Existing techniques for forming collaborative groups emphasize on effective transfer of knowledge between learners. However, mastery of any subject requires not just knowledge, but a distinct set of learning abilities. While forming collaborations, one must examine the scope for mutual transfer of abilities that different members possess so as to fulfill any skills gap. In this paper, we illustrate two well-known algorithms: the Stable Marriages Problem and the Kernighan-Lin algorithm to discover optimal competitive and complementary groupings between students. Experiments reveal that competitive (complementary) collaborative pairs generated by either approach yield the same pairs when the participants have matching (contrasting) performance in 50% or more parameters. The Stable Marriages approach takes lesser time to discover all possible pairs while K-Lin has the distinct advantage of generating multi-member groups.