AnonyMED-BR is a dataset created for research on medical text anonymization in Brazilian Portuguese.It combines real electronic health records (EHRs) — used strictly for research under ethics committee approval — with synthetically generated medical records to support the development of robust transformer-based models.
Language(s): Brazilian Portuguese (pt-BR)
Domain: Clinical / Medical
Task(s): Named Entity Recognition (NER)… See the full description on the dataset page: https://huggingface.co/datasets/Venturus/AnonyMED-BR.
