metadata
license: unknown
task_categories:
- token-classification
language:
- de
tags:
- medical
size_categories:
- 1K<n<10K
There is also a version with classLabels here: 2ndBestKiller/DrugTestWithClassLabels
The set consists of a mix of german wikipedia articles, medical guidelines, medication analyses, patient informations and AI generated text. The dataset has been auto annotated with a phrase matcher (exact matching) with a list of all avaivable medical substances in germany as listed by the BfArM (though it does not include brand names). Annotations have not been checked manually!