EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Early Detection of Cognitive Fatigue in E-Learning Environments Using Explainable AI and Multimodal Sensor Data

Bayda Zahid Kamil, Ohood Fadil Alwan, Zainab Yaseen Taha Taqa

International Journal on Advanced Science, Engineering and Information Technology · 17.10.2026 · 16 (5) · 1643-1647

DOI: 10.18517/ijaseit.16.5.14571 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal on Advanced Science, Engineering and Information Technology, online erschienen am 17.10.2026. Verfasst von Bayda Zahid Kamil und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

As e-learning environments have grown rapidly, cognitive fatigue has become a crucial factor that harms learners' attention, performance, and engagement. To address this issue, the present study proposes an intelligent framework for cognitive fatigue detection through multimodal sensor fusion using Explainable Artificial Intelligence (XAI). We evaluated the model's performance on multimodal cohort data from the CogBeacon dataset, which includes EEG signal features, facial feature volumes, physiological attributes, and self-reported fatigue scores. We built and compared three machine learning classifiers: Random Forest (RF), Support Vector Machine (SVM), and Multilayer Perceptron (MLP). The following table details the prediction accuracy, recall, and F1-score for all models on the test data. The experimental results showed that Random Forest was the best-performing model overall, with 85% accuracy, followed by AdaBoost, which performed second best with 83% recall and an F1-score of 84%. Additionally, SHAP explainability analysis showed EEG alpha and theta activity, heart rate variability, and blink frequency as the four most important cognitive fatigue predictors. The results of this work on the feasibility of sensor-based monitoring technologies integrated with XAI approaches for fatigue estimation can be applied to provide adapted and trustworthy feedback mechanisms to support adaptive interventions in e-learning systems. AI-driven solutions like these can support cost-effective, scalable, retention-boosting student learning.

Zitieren

APA

Kamil, B. Z., Alwan, O. F., & Taqa, Z. Y. T. (2026). Early Detection of Cognitive Fatigue in E-Learning Environments Using Explainable AI and Multimodal Sensor Data. International Journal on Advanced Science, Engineering and Information Technology, 16(5), 1643-1647. https://doi.org/10.18517/ijaseit.16.5.14571

Vancouver

Kamil BZ, Alwan OF, Taqa ZYT. Early Detection of Cognitive Fatigue in E-Learning Environments Using Explainable AI and Multimodal Sensor Data. International Journal on Advanced Science, Engineering and Information Technology. 2026;16(5):1643-1647. doi:10.18517/ijaseit.16.5.14571

BibTeX
@article{Kamil2026,
  author = {Kamil, Bayda Zahid and Alwan, Ohood Fadil and Taqa, Zainab Yaseen Taha},
  title = {Early Detection of Cognitive Fatigue in E-Learning Environments Using Explainable AI and Multimodal Sensor Data},
  journal = {International Journal on Advanced Science, Engineering and Information Technology},
  year = {2026},
  volume = {16},
  number = {5},
  pages = {1643--1647},
  doi = {10.18517/ijaseit.16.5.14571}
}
Kostenlos starten →