온톨로지 기반 의미 라벨링을 이용한 EEG 인지 상태 클러스터링 프레임워크 = Ontology Based Semantic Labeling Framework for EEG Cognitive State Clustering

다국어 초록 (Multilingual Abstract)

This paper proposes an ontology-based semantic labeling framework to address the interpretability limitation of unsupervised clustering results in EEG-based cognitive state analysis.
Conventional approaches mainly rely on supervised classification, which requires extensive labeled data and suffers from limited generalization across users. To overcome these limitations, the proposed framework applies unsupervised clustering to EEG feature representations and assigns semantic cognitive state labels using an ontology-based knowledge model. The framework separates data-driven analysis from knowledge-driven interpretation while integrating them at the semantic labeling stage.
Experimental analysis demonstrates that clustered EEG patterns can be systematically mapped to interpretable cognitive state concepts through ontology-based rules. The proposed framework enhances the interpretability of EEG cognitive state analysis and can be flexibly applied to various IoT-based intelligent service environments.