Unlabeled Parsing on Penn Treebank WSJ10 (test)
82.9F1 (max)UML-DOP
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| UML-DOPTraining Data=WSJ10 Full2018.10 | 82.9 | — | — | |
| DMV+CCMTraining Data=WSJ10 Full2018.10 | 77.6 | — | — | |
| CCMTraining Data=WSJ10 Full2018.10 | 71.9 | — | — | |
| PRPN-LMTraining Data=WSJ Train, Training Object=LM, Vocab Size=10k2018.10 | 71.3 | 70.5 | 0.4 | |
| PRPN-UPTraining Data=WSJ Train, Training Object=LM, Vocab Size=15.8k2018.10 | 70.3 | 62.2 | 3.9 | |
| PRPN-UPTraining Data=AllNLI Train, Training Object=LM, Vocab Size=76k2018.10 | 68.5 | 66.3 | 0.8 | |
| ON-LSTM 2nd-layerTraining Data=WSJ Train, Training Object=LM, Vocab Size=10k2018.10 | 66.8 | 65.1 | 1.7 | |
| PRPN-LMTraining Data=AllNLI Train, Training Object=LM, Vocab Size=76k2018.10 | 58.1 | 52.4 | 4.9 | |
| ON-LSTM 3rd-layerTraining Data=WSJ Train, Training Object=LM, Vocab Size=10k2018.10 | 57.6 | 54 | 3.9 | |
| Right Branching2018.10 | 56.6 | 56.6 | 0 | |
| Balanced Trees2018.10 | 43.4 | 43.4 | 0 | |
| ON-LSTM 1st-layerTraining Data=WSJ Train, Training Object=LM, Vocab Size=10k2018.10 | 42.8 | 35.2 | 4.1 | |
| Random Trees2018.10 | 32.2 | 31.7 | 0.3 | |
| Left Branching2018.10 | 19.6 | 19.6 | 0 |