Predicting students' success based on forum activities in MOOCs

Massive Open Online Courses (MOOCs) have the potential to scale university education, allowing for many thousands of students to participate in a single online course. But even successful MOOC platforms like Coursera, edX or Iversity face the problem of very low completion rates. Analyzing learning activities in a MOOC, learning analytics could help to identify necessary interventions. In Iversity MOOCs, a forum is the basic communication platform, both for student-student and student-instructor communication. In this paper, features of successful students are derived from forum activities and combined to a learning profile. From this learning profile, feedback could be generated for students who are classified as ”risk students”. An analytics tool was developed, based on machine learning, which classifies students in Iversity MOOCs, using features like number of answers in a forum or number of up-votes. Thus, features of successful students can be determined, and visualized in an intuitive way.