Table 20 Results of constituent parsers using their distributed non-biomedical models on the CRAFT development set; labeled bracket precision (LB-P), recall (LB-R) and F-score (LB-F)
Parser | LB-P | LB-R | LB-F | unevaluated count |
|---|---|---|---|---|
Berkeley | 61.60 | 64.50 | 63.02 | 4 |
Bikel | 63.97 | 65.82 | 64.89 | 2 |
Charniak-Johnson | 62.51 | 65.55 | 64.00 | 59 |
Enju | 71.93 | 43.56 | 54.26 | 8 |
Mogura | 54.74 | 43.25 | 48.32 | 8 |
Stanford 1.6 | 60.76 | 64.70 | 62.67 | 3 |