Table 5. Summary table with performance metrics reported as median (95% CI) and F 1 interquartile variance (IQV) after 200 bootstrap iterations.
The performance metrics are compared across pipelines using different distributed document representations.
| Pipeline | Precision (%) | Recall (%) | F 1 (%) | F 1 IQV |
|---|---|---|---|---|
| SPECTER | 74.1 (66.5,80.9) | 69.0 (62.0,76.6) | 71.2 (66.2,75.8) | 9.6 |
| BioBERT mean pooling | 78.1 (69.0,85.4) | 75.3 (68.3,82.9) | 76.6 (71.6,81.3) | 9.7 |
| BioBERT mean + min&max pooling | 80.1 (71.8,86.0) | 75.9 (69.6,82.9) | 77.7 (72.7,81.4) | 8.7 |