Table 10.

Binary relation detection results for both EEG decoding components for the relation types Job Title and Visited using GloVe embeddings.

	Recurrent EEG decoding			Convolutional EEG decoding
Job title vs. None	Precision	Recall	F1-score	Precision	Recall	F1-score
GloVe	0.789	0.776	0.767 (0.05)	0.789	0.776	0.767 (0.05)
GloVe + EEG full	0.792	0.782	0.773 (0.06)	0.796	0.793	0.789 (0.05)
GloVe + EEG γ	0.780	0.788	0.774 (0.1)	0.817	0.811	0.808 (0.03)
Visited vs. None	Precision	Recall	F1-score	Precision	Recall	F1-score
GloVe	0.762	0.756	0.734 (0.1)	0.762	0.756	0.734 (0.1)
GloVe + EEG full	0.756	0.759	0.745 (0.1)	0.766	0.758	0.750 (0.09)
GloVe + EEG γ	0.773	0.768	0.754 (0.1)	0.819	0.795	0.795 (0.09)

The best result in each column is marked in bold.