SiMoNERo

Romanian biomedical Named Entity Recognition dataset, converted to HuggingFace format.

Entity types: ANAT (body parts), CHEM (chemicals & drugs), DISO (disorders), PROC (procedures).

Splits

Split	Sentences
train	3,747
validation	443
test	491

NER Tags

ID	Tag
0	O
1	B-ANAT
2	B-CHEM
3	B-DISO
4	B-PROC
5	I-ANAT
6	I-CHEM
7	I-DISO
8	I-PROC

Usage

from datasets import load_dataset

ds = load_dataset("xd-br0/SiMoNERo")

print(ds["train"][0])
print(ds["train"].features["ner_tags"].feature.names)
# ['O', 'B-ANAT', 'B-CHEM', 'B-DISO', 'B-PROC', 'I-ANAT', 'I-CHEM', 'I-DISO', 'I-PROC']

Citation

@inproceedings{mititelu-mitrofan-2020-simonero,
    title     = "The {R}omanian Medical Treebank -- {SiMoNERo}",
    author    = "Mititelu, Verginica Barbu and Mitrofan, Maria",
    booktitle = "Proceedings of the 15th International Conference on Linguistic
                 Resources and Tools for Natural Language Processing (ConsILR-2020)",
    year      = "2020",
    pages     = "7--16",
    ISSN      = "1843-911X"
}

@inproceedings{mitrofan-etal-2019-monero,
    title     = "{M}o{NER}o: a Biomedical Gold Standard Corpus for the {R}omanian Language",
    author    = "Mitrofan, Maria and Barbu Mititelu, Verginica and Mitrofan, Grigorina",
    booktitle = "Proceedings of the 18th BioNLP Workshop and Shared Task",
    month     = aug,
    year      = "2019",
    address   = "Florence, Italy",
    publisher = "Association for Computational Linguistics",
    url       = "https://aclanthology.org/W19-5008/",
    pages     = "71--79"
}

@inproceedings{mitrofan-pais-2022-improving,
    title     = "Improving {R}omanian {B}io{NER} Using a Biologically Inspired System",
    author    = "Mitrofan, Maria and Pais, Vasile",
    booktitle = "Proceedings of the 21st Workshop on Biomedical Language Processing",
    month     = may,
    year      = "2022",
    address   = "Dublin, Ireland",
    publisher = "Association for Computational Linguistics",
    url       = "https://aclanthology.org/2022.bionlp-1.30/",
    doi       = "10.18653/v1/2022.bionlp-1.30",
    pages     = "316--322"
}