Detecting Internal Distraction in an Educational VR Environment Using EEG Data

Abstract

Virtual Reality (VR) provides a more engaging learning experience to students and could improve knowledge retention compared to traditional learning methods. However, a student could get distracted in the VR environment due to stress, mind-wandering, unwanted noise, external sounds, etc. Distractions could be classified as either external (due to environment) or internal (due to internal thoughts). Past researchers have used eye-gaze data to detect external distractions. However, eye-gaze data can not measure internal distractions since a user could look at the educational content and may be thinking about something else. We explored the usage of electroencephalogram (EEG) data to detect internal distraction. We designed an educational VR environment and trained three machine learning models: Random Forest (RF), Support Vector Machine (SVM), and Linear Discriminant Analysis (LDA), to detect internal distractions of students. Our preliminary study results show that RF provides a better accuracy (98%) compared to SVM and LDA.