Skip to main content

Table 7 Evaluation results from our system using SA weighting on the BMC dataset

From: Recognizing speculative language in biomedical research articles: a linguistically motivated perspective

Threshold Precision Recall Accuracy F1 score
1 0.58 0.96 0.87 0.73
2 0.66 0.94 0.90 0.77
3 0.80 0.85 0.94 0.82
4 0.83 0.65 0.92 0.73
5 0.95 0.56 0.92 0.70
6 0.97 0.35 0.89 0.52
7 0.98 0.21 0.86 0.35