Course recommendation systems that learn from course and learner attributes can be important tools for course discovery in a rapidly evolving course catalog. In this talk, we will present methodology and results from content-based and learner-based recommendation systems that we built and deployed for Open edX courses, using course and learner data. We will cover challenges faced in implementing such systems with the available Open edX data and hope to have a discussion on how such systems could be extended, or could inform data formats and course analytics strategies.
