Skip to main content
[Preprint]. 2024 Mar 11:rs.3.rs-4014472. [Version 1] doi: 10.21203/rs.3.rs-4014472/v1

Table 2.

Performance of the rule-based NLP tools at the three sites

WCM NM UF

Suicidal History
 Accuracy 0.82 0.79 0.84
 Precision (macro average) 0.81 0.67 0.77
 Recall (macro average) 0.83 0.69 0.60
 F1-score (macro average) 0.81 0.68 0.63
Family Suicidal History
 Accuracy 0.94 0.91 0.98
 Precision (macro average) 0.79 0.85 0.99
 Recall (macro average) 0.80 0.44 0.62
 F1-score (macro average) 0.80 0.58 0.69