Unsupervised extraction of students navigation patterns on an EPFL MOOC

How do students learn in MOOCs? This project aims at answering this question by analyzing the activities of thousands of students registered on EPFL Scalaa MOOC hosted by Coursera. With the rapid growth of MOOCs, Education Science has entered the Big Data bubble, bringing new opportunities to study and improve learning technologies. We are interested in studying students navigation patterns which are the short sequences of learning activities that a students perform on the MOOC platform. In our case, the learning activities are one of watching a video lecture, reading or posting on the forum and submitting assignments. In this project we use unsupervised machine learning techniques to extract the main navigation patterns of students and gain insights on their behavior. We produce a simple and efficient visualization tool in order to provide feedback to teachers to help them understand the potential difficulties encountered by their students during the course and, if necessary, take actions accordingly