| E. Dritsas, M. Trigka, Ph. Mylonas |
| Mental Confusion Prediction in e-Learning Contexts with EEG and Machine Learning |
| Novel & Intelligent Digital Systems Conference (NiDS 2023), September 28-29, 2023, Athens, Greece |
|
ABSTRACT
|
| In cognitive science, the term confusion is used to capture the decline in learnersą cognitive ability, which affects their ability to think, solve a problem, learn and understand. Unlike classroom education, in online e-learning contexts (such as in Massive Open Online Courses (MOOC)), confusion hinders the smooth evolution of the learning process from the learnersą side, as the educator canąt immediately interact with the students in order to restore cognitive equilibrium. This paper presents a Machine Learning (ML) based approach by comparing several classifiers that were trained and tested exploiting Electroencephalogram (EEG) data (namely, band power, attention and mediation features) acquired by the MindSet device in order to efficiently distinguish “Confused” from “Not-Confused” subjects. In particular, the J48 was the dominant model reaching an optimal performance with accuracy, precision and recall equal to 99.9%, and an Area Under the Curve (AUC) of 100%.
|
| 28 September, 2023 |
| E. Dritsas, M. Trigka, Ph. Mylonas, "Mental Confusion Prediction in e-Learning Contexts with EEG and Machine Learning", Novel & Intelligent Digital Systems Conference (NiDS 2023), September 28-29, 2023, Athens, Greece |
[  PDF] [
BibTex] [
Print] [
 Back] |
