Diagnose Topic Attention: What are the Preference and Demand with Different Learner Groups in MOOCs Discussion Forums

Abstract

The autonomy and openness of MOOC provide a lifelong learning channel for many learners. As an asynchronous interactive community, the discussion forum has accumulated a large amount of behaviors and text dataset, which can reflect learners' academic information. Learning preferences and learning demands are very important. They are not only the starting point of instructional design, but also the basis for teachers and managers to design intervention schemes and optimize teaching. However, the current diagnosis methods have the disadvantages of high data acquisition cost and long sampling period, and cannot give timely and accurate feedback to the diverse preferences and demands of different learning groups. Therefore, this paper presents a new method, which is based on the TEAM model and combined with the data characteristics with content and behavior in the MOOC discussion forum, we calculate learners' topic attention and diagnose the preferences and demands of learning groups with or without certificates accurately. The results show that in the whole teaching cycle, all learners' knowledge demands and preference appear concomitantly, and the learner groups with certificates show a strong willingness to task-based demands and preferences.