A DISCOURSE CENTRIC FRAMEWORK FOR FACILITATING INSTRUCTOR INTERVENTION IN MOOC DISCUSSION FORUMS

Abstract

We propose a discourse framework - predictive models that use discourse level information in student posts - to guide Massive Open Online Course (MOOC) instructors to selectively post on student discussions on forums, here- after referred to as interventions, which otherwise is infeasible to scale. Prior works on intervention guidance, from the pre-MOOC era, and on MOOCs do not address diversity and scale. Our models and evaluation explicitly cater to diversity and scale, both inherent to MOOCs. Simplistic vocabulary-based prediction models fail to adapt sufficiently due to the diversity in MOOCs. To address this we use discourse relations, that are not only domain-agnostic but also signal intervention. This further leads us to investigate inter-post discourse structures. We propose a pedagogically grounded discourse taxonomy and build an annotated corpus of student posts in instructor intervened threads. We propose a debiasing classifier to unlearn the bias and predict interventions. We show that neural model of thread as a sequence of posts significantly improve the state-of-the-art towards production- ready models. Our models can be integrated to an instructor dashboard that can flag a discussion thread.