Table 3.
Simple baseline - Bag-of-Words (BoW) Model | ||||||
---|---|---|---|---|---|---|
Labels | Stanford dataset | UPMC dataset | ||||
Precision | Recall | F1-score | Precision | Recall | F1-score | |
PE acute | 72.38% | 83.53% | 77.11% | 82.04% | 81.55% | 81.79% |
PE present | 78.15% | 88.75% | 82.49% | 83.19% | 83.46% | 83.33% |
PE subsegmental only | 72.82% | 76.92% | 74.70% | N/A |