Table 6.
Relation detection results of the multi-modal model using the recurrent EEG decoding component.
| Randomly initialized | GloVe | BERT | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Model | P | R | F1 (std) | P | R | F1 (std) | P | R | F1 (std) |
| Baseline | 0.404 | 0.525 | 0.452 (0.04) | 0.501 | 0.609 | 0.539 (0.05) | 0.522 | 0.788 | 0.623 (0.05) |
| + noise | 0.420 | 0.424 | 0.408 (0.07) | 0.577 | 0.497 | 0.532 (0.03) | 0.675 | 0.585 | 0.625 (0.03) |
| + ET | 0.421 | 0.404 | 0.402 (0.06) | 0.547 | 0.476 | 0.506 (0.04) | 0.661 | 0.631 | 0.644 (0.03) |
| + EEG full | 0.345 | 0.343 | 0.334 (0.05) | 0.511 | 0.387 | 0.432 (0.09) | 0.652 | 0.690 | 0.668* (0.10) |
| + EEG θ | 0.430 | 0.421 | 0.414 (0.07) | 0.582 | 0.508 | 0.539 (0.07) | 0.646 | 0.736 | 0.684* (0.08) |
| + EEG α | 0.368 | 0.373 | 0.358 (0.12) | 0.582 | 0.515 | 0.542 (0.06) | 0.652 | 0.715 | 0.679* (0.07) |
| + EEG β | 0.349 | 0.340 | 0.329 (0.09) | 0.581 | 0.497 | 0.532 (0.10) | 0.674 | 0.726 | 0.696+ (0.06) |
| + EEG γ | 0.410 | 0.399 | 0.397 (0.05) | 0.554 | 0.488 | 0.514 (0.09) | 0.666. | 0.715 | 0.686* (0.07) |
| +θ+α+β+γ | 0.370 | 0.376 | 0.363 (0.09) | 0.554 | 0.488 | 0.514 (0.09) | 0.675 | 0.646 | 0.659 (0.04) |
We report precision (P), recall (R), F1-score and the standard deviation (std) between five runs. The best results per column are marked in bold. Significance is indicated on the F1-score with asterisks:
*
denotes p < 0.05 (uncorrected), + denotes p < 0.003 (Bonferroni corrected p-value).