Skip to Main Content
Table 4

Held-out results for identifying speculation cues: Applying the cue classifiers to the 5,003 sentences in BSE — the biomedical papers provided for the CoNLL-2010 Shared Task evaluation.


Sentence Level
Token Level
Cue Level
Model
Prec
Rec
F1
Prec
Rec
F1
Prec
Rec
F1
Baseline 77.59 81.52 79.51 77.16 72.39 74.70 75.15 72.49 73.80 
WbW 89.28 83.29 86.18 87.62 73.95 80.21 86.33 74.21 79.82 
Filtering 87.87 85.32 86.58 86.46 76.74 81.31 84.79 77.17 80.80 
Tang et al. 2010 85.03 87.72 86.36 n/a n/a n/a 81.70 80.99 81.34 

Sentence Level
Token Level
Cue Level
Model
Prec
Rec
F1
Prec
Rec
F1
Prec
Rec
F1
Baseline 77.59 81.52 79.51 77.16 72.39 74.70 75.15 72.49 73.80 
WbW 89.28 83.29 86.18 87.62 73.95 80.21 86.33 74.21 79.82 
Filtering 87.87 85.32 86.58 86.46 76.74 81.31 84.79 77.17 80.80 
Tang et al. 2010 85.03 87.72 86.36 n/a n/a n/a 81.70 80.99 81.34 
Close Modal

or Create an Account

Close Modal
Close Modal