Transition-based transducer
Reported on 8 benchmarks across 2 tasks
Note: results are matched by exact model name. Different papers may use the same name for different model variants.
Natural Language Processing8 results
- F1 Full66best: 78.5 (APT (IBM))
- F1 Newswire70best: 73.3 (Transition-based+improved aligner+ensemble)
- F1 Full0.66best: 0.7 (Sequence-to-Graph Transduction)
- F1 Newswire0.7best: 0.75 (Sequence-to-Graph Transduction)
- F1 Full66best: 78.5 (APT (IBM))
- F1 Newswire70best: 73.3 (Transition-based+improved aligner+ensemble)
- F1 Full0.66best: 0.7 (Sequence-to-Graph Transduction)
- F1 Newswire0.7best: 0.75 (Sequence-to-Graph Transduction)