Deciphering the attributes of student retention in massive open online courses using data mining techniques

Abstract

Aimed at a massive outreach and open access education, Massive Open Online Courses (MOOC) has evolved incredibly engaging millions of learners’ over the years. These courses provide an opportunity for learning analytics with respect to the diversity in learning activity. Inspite of its growth, high dropout rate of the learners’, it is examined to be a paramount factor that may obstruct the development of the e-learning platforms. Fabricating on the existing efforts of retaining learners’ engagement prior to learning, the study explores to decipher the attributes of student retention in e- learning. The study proposes a clear rationale of significant attributes using classification algorithms (Decision Tree) in order to improve course design and delivery for different MOOC providers and learners’. Using the three MOOC datasets, this research work analyses the approach and results of applying the data mining techniques to online learners’, based on their in-course behaviour. Finally, it predicts the attributes that lead to minimise attrition rate and analyse the different cohort behaviour and its impacts for dropouts using data mining technique. It focuses to build a more integrated environment for these learners’.