An Unsupervised Method for Discovering How Does Learners Progress toward Understanding in MOOCs

Massive Open Online Course (MOOC) seems to expand access education and it present too many advantages as: democratization of learning, openness all accessibility on a large scale, etc. However, this new phenomenon open learning suffers from the lack personalization; is not easy identify learners’ characteristics because their heterogeneous masse. Following increasing adoption styles as personalization criteria, possible make process easier for learners. In paper, we extracted features traces when they interact with MOOC platform in order an automatic way. For purpose, adopted Felder-Silverman Learning Style Model (FSLSM) used unsupervised clustering method. Finally, solution was implemented clustered learners based level preference sequential/global dimension FSLSM. Results indicated that, first: k-means best performing algorithm comes identification styles; second: majority show strong moderate sequential style preferences.