TALE 2021: Topic Modeling in MOOCs: What Was to Be Discussed, What the Instructor Discussed, and What the Learners Discussed

This is another paper that had a long journey. We first tested the techniques on a different MOOC (Modern Japanese Architecture) before testing on the Autophagy MOOC.

Zoom virtual background

For a copy of relevant materials (e.g., presentation, paper) or any questions you may have, please feel free to reach out to me through the Contact Me gadget on this blog's sidebar.

Details

Title: Topic Modeling in MOOCs: What Was to Be Discussed, What the Instructor Discussed, and What the Learners Discussed

Authors: May Kristine Jonson Carlon, Anie Day DC. Asa, Nopphon Keerativoranan, Toru Nagahama, Jeffrey S. Cross

Venue: IEEE International Conference on Engineering, Technology, and Education TALE 2021

Date: December 5 to 8, 2021

DOI: 10.1109/TALE52509.2021.9678621 ($$)

Abstract

Instructors create a course outline to serve as a guide for content delivery throughout the course. While contents can be adjusted based on learner feedback and performance in instructor-paced courses, the same may not be said for massive open online courses (MOOCs) that run continuously for months or years. For example, instructors cannot make drastic changes to self-paced MOOC content until the course is closed, revised, and re-released. Topic modeling was done on a microbiology MOOC to see if it could be a viable means of assessing how well the  MOOC  content was able to capture the course outline concepts. The same approach was applied to discussion board posts to identify which topics resonated with the learners. Topic modeling can be useful in achieving the said goals. This can be impactful in introducing improvements to the online course even before the course is revised and re-released.