F1 scores for distant supervision performance, evaluated on the OIAM training set, with different sets of stopwords, and training on either CKS topics or ICPC-2 descriptions
| Model | No removal | English | Medical | Custom | Medical+custom | English+custom | English+medical+custom |
| NB (multilabel), ICPC-2 | 0.139 | 0.170 | 0.136 | 0.253 | 0.297 | 0.323 | 0.297 |
| NB (multilabel), CKS | 0.096 | 0.160 | 0.119 | 0.126 | 0.191 | 0.207 | 0.234 |
| NB (multiclass), ICPC-2 | 0.324 | 0.354 | 0.307 | 0.461 | 0.471 | 0.512 | 0.470 |
| NB (multiclass), CKS | 0.245 | 0.274 | 0.249 | 0.275 | 0.340 | 0.368 | 0.375 |
| Nearest centroid, ICPC-2 | 0.312 | 0.354 | 0.317 | 0.432 | 0.437 | 0.445 | 0.437 |
| Nearest centroid, CKS | 0.326 | 0.349 | 0.344 | 0.349 | 0.353 | 0.357 | 0.365 |
CKS, Clinical Knowledge Summaries; ICPC-2, International Classification of Primary Care-2; NB, Naïve Bayes; OIAM, One in a Million.