Modeling Students' Attention in the Classroom Using Eyetrackers
ACM Southeast Conference- ACMSE 2019- Session 1: Long Papers
The process of learning is not merely determined by what the instructor teaches, but also by how the student receives that information. An attentive student will naturally be more open to obtaining knowledge than a bored or frustrated student. In recent years, tools such as skin temperature measurements and body posture calculations have been developed for the purpose of determining a student's affect, or emotional state of mind. However, measuring eye-gaze data is particularly noteworthy in that it can collect measurements non-intrusively, while also being relatively simple to set up and use. This paper details how data obtained from an eye-tracker can indeed be used to predict a student's attention as a measure of affect over the course of a class. From this research, an accuracy of 77% was achieved using the Extreme Gradient Boosting technique of machine learning. The outcome indicates that eye-gaze can be indeed used as a basis for constructing a predictive model.
Veliyath, Narayanan, Pradipta De, Andrew A. Allen, Charles B. Hodges, Anirudda Mitra.
"Modeling Students' Attention in the Classroom Using Eyetrackers."
ACM Southeast Conference- ACMSE 2019- Session 1: Long Papers Kennesaw, GA.
doi: 10.1145/3299815.3314424 source: https://dl.acm.org/citation.cfm?doid=3299815.3314424 isbn: 978-1-4503-6251-1