Table 4 Performance using different parsers on the development set.
From: Optimizing graph-based patterns to extract biomedical events from the literature
Parser Type | Event pattern | Recall | Precision | F-score |
|---|---|---|---|---|
Charniak | 2,923 | 47.01% | 66.01% | 54.91% |
Stanford | 3,305 | 43.66% | 67.67% | 53.08% |
Ensemble | 4,617 | 47.45% | 65.65% | 55.09% |