Leveraging Emotions in Student Feedback to Improve Course Content and Delivery

Abstract

Emotions play a vital role in almost all the activities we perform, including learning. In fact, the success of any learning system is largely dependent upon its ability to deliver the course content in such a form so as to meet the learning requirements of the target audience. Learning Systems can be tailored to effectively utilize the feedback from learners to improve the course content, and thus the feedback can prove to be a valuable asset. There is an increased demand for focusing on a learner-centric approach to content delivery. In this study we attempt at detecting different learning-relevant emotions from the feedback for a course so as to enable course designers to incorporate the type of content that matches a learners requirements. Rather than taking into account six basic emotions (sadness, happiness, fear, anger, surprise and disgust) we consider interest, engagement, confusion, frustration, disappointment, boredom, hopefulness and satisfaction emotions for the purpose of our study since they are more relevant in a learning setup. We employed a supervised algorithm, Support Vector Machine, for affect detection from the textual feedback in our experiments.