Datasets:

Languages:
English
License:

The viewer is disabled because this dataset repo requires arbitrary Python code execution. Please consider removing the loading script and relying on automated data support (you can use convert_to_parquet from the datasets library). If this is not possible, please open a discussion for direct help.

Dataset Card for BEAR

A dataset of 2100 Twitter posts annotated with 14 different types of biomedical entities (e.g., disease, treatment, risk factor, etc.) and 20 relation types (including caused, treated, worsens, etc.).

Citation Information

@InProceedings{wuehrl_klinger_2022,
  author    = {Wuehrl, Amelie  and  Klinger, Roman},
  title     = {Recovering Patient Journeys: A Corpus of Biomedical Entities and Relations on  Twitter (BEAR)},
  booktitle      = {Proceedings of The 13th Language Resources and Evaluation Conference},
  month          = {June},
  year           = {2022},
  address        = {Marseille, France},
  publisher      = {European Language Resources Association}
}
Downloads last month
44