Skip to main content
. 2021 Aug 19;4(3):ooab071. doi: 10.1093/jamiaopen/ooab071

Table 4.

The comparison of our proposed approach to existing baseline models for identification of gold standard medication mentions in our evaluation set

Model No. of true positives No. of false positives No. of false negatives Precision (%) Recall (%) F-score (%)
cTAKES 1119 2814 163 28.5 87.3 42.9
MedEx-UIMA 830 1215 292 40.6 74.0 52.4
MedXN 832 318 432 72.3 65.8 68.9
Our approach 1062 168 206 86.3 83.8 85.0