Publication

Abstract : Massive Open Online Courses (MOOCs) have gained immense popularity in the last few years. An integral part of any course that runs on a MOOC platform is the learners' interaction. Most of this interaction happens today via forums that are built into the course section. In this paper, we begin by highlighting the purpose of forums in MOOCs and its shortcomings such as quantity of participation, good set of resident or visiting experts, unregulated exchange of information and lack of auto answering. Then we turn our focus on analysing the data of successful Community Question and Answer (CQamp;A) sites such as Stack Overflow, Quora and Yahoo! Answers using data mining techniques. Based on what we learnt, we list features that are needed in MOOC forums. Finally, we also propose a forum footprint that can adjust dynamically based on community behaviour.