Rhetorical Role Labeling on DEEPRHOLE
47.9Macro-F1Gold Prototypes
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Gold PrototypesEvaluation protocol=optimal prototype post-hoc for each sentence2026.03 | 47.9 | 56.02 | |
| PCMSampling strategy=Random Sampling2026.03 | 47.3 | 53.9 | |
| PCMSampling strategy=Full Sampling2026.03 | 47.13 | 55.54 | |
| PCMSampling strategy=Supervised Sampling2026.03 | 45.92 | 53.86 | |
| Mind2026.03 | 45.3 | 50.93 | |
| PBR2026.03 | 44.96 | 51.11 | |
| Baseline (HSLN)2026.03 | 44.24 | 50.51 |