본문 바로가기
  • Home

A Personalized Learning Recommendation Method Using SOM-Based BCI User State Classification

  • Journal of Internet of Things and Convergence
  • Abbr : JKIOTS
  • 2025, 11(6), 5
  • Publisher : The Korea Internet of Things Society
  • Research Area : Engineering > Computer Science > Internet Information Processing
  • Received : October 27, 2025
  • Accepted : December 10, 2025
  • Published : December 31, 2025

Guijung Kim 1

1백석대학교

Accredited

ABSTRACT

The proposed system demonstrates the feasibility of adaptive learning environments that respond to learners’ cognitive and physiological changes in real time. This study proposes an unsupervised learning approach using the Self-Organizing Map (SOM) to classify user states in Brain– Computer Interface (BCI) systems and apply them to a personalized learning recommendation model. EEG signals from alpha, beta, and theta frequency bands were analyzed to cluster user cognitive states into high focus, moderate focus, and low focus levels. The SOM model enabled clustering without labeled data, and the ontology-based recommendation algorithm dynamically adjusted learning content according to user states. Future work will focus on real EEG experiments and the integration of multimodal biosignals to enhance precision and expand the practical application of real-time personalized learning systems.

Citation status

* References for papers published after 2024 are currently being built.