Discourse Parsing (with gold EDU segmentation) on RST-DT (test)
97.44Span ScoreLin et al. (2019)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Lin et al. (2019)training strategy=joint2021.06 | 97.44 | 91.34 | 81.7 | |
| Conditional Splitting Frameworksegmentation setting=gold EDU2021.06 | 97.37 | 91.95 | 82.1 | |
| Lin et al. (2019)training strategy=single2021.06 | 96.94 | 90.89 | 81.28 | |
| Human Agreement2021.06 | 95.7 | 90.4 | 83 | |
| Wang et al. (2017)2021.06 | 95.6 | 87.8 | 77.6 |