Analysing and Predicting Recurrent Interactions among Learners during Online Discussions in a MOOC

High attrition rates are one of the biggest concerns in MOOCs. One of the possible causes may be learners’ lack of interactions and low levels of participations in MOOCs online discussions. Research to measure and predict recurrent interactions of learners in MOOCs online discussions has the potential to gain inside into the likely impact on the attrition rate. It is argued that personalisation in MOOCs has the potential to increase learners’ interactions and associated factors to continuous friendships. In this paper, a detailed analysis has been carried out of learners’ interactions within a MOOC. This paper investigates learners’ interaction habits and their recurrent interactions throughout the entire duration of a MOOC’s course, and consequently proposes a method to measure the interactions and predict possible interactions between peers. The findings denote that when a learner interacted with their peer, they most probably interact again in the following weeks. Moreover, our proposed prediction method also demonstrate promising results towards predicting future interactions between learners based on their previous relationships.