README.md

---
language:
- en
- fr
- gal
- it
- pt
- ro
- cat
multilinguality:
- multilingual
license: cc-by-4.0
task_categories:
- token-classification
pretty_name: LivingNER
config_names:
- en
- fr
- gal
- it
- pt
- to
- cat
- combined

dataset_info:
- config_name: en
  splits:
  - name: train
    num_bytes: 14699476
    num_examples: 1000
  - name: validation
    num_bytes: 6764942
    num_examples: 500
- config_name: fr
  splits:
  - name: train
    num_bytes: 14699476
    num_examples: 1000
  - name: validation
    num_bytes: 6764942
    num_examples: 500
- config_name: gal
  splits:
  - name: train
    num_bytes: 14699476
    num_examples: 1000
  - name: validation
    num_bytes: 6764942
    num_examples: 500
- config_name: it
  splits:
  - name: train
    num_bytes: 14699476
    num_examples: 1000
  - name: validation
    num_bytes: 6764942
    num_examples: 500
- config_name: pt
  splits:
  - name: train
    num_bytes: 14699476
    num_examples: 1000
  - name: validation
    num_bytes: 6764942
    num_examples: 500
- config_name: ro
  splits:
  - name: train
    num_bytes: 14699476
    num_examples: 1000
  - name: validation
    num_bytes: 6764942
    num_examples: 500
- config_name: cat
  splits:
  - name: train
    num_bytes: 14699476
    num_examples: 1000
  - name: validation
    num_bytes: 6764942
    num_examples: 500
- config_name: combined
  splits:
  - name: train
    num_bytes: 108745150,
    num_examples: 7000
  - name: validation
    num_bytes: 50100231
    num_examples: 3500
    
configs:
- config_name: default
  data_files:
  - split: train
    path: en/train/data-*
  - split: validation
    path: en/validation/data-*
- config_name: en
  data_files:
  - split: train
    path: en/train/data-*
  - split: validation
    path: en/validation/data-*
- config_name: fr
  data_files:
  - split: train
    path: fr/train/data-*
  - split: validation
    path: fr/validation/data-*
- config_name: gal
  data_files:
  - split: train
    path: gal/train/data-*
  - split: validation
    path: gal/validation/data-*
- config_name: it
  data_files:
  - split: train
    path: it/train/data-*
  - split: validation
    path: it/validation/data-*
- config_name: pt
  data_files:
  - split: train
    path: pt/train/data-*
  - split: validation
    path: pt/validation/data-*
- config_name: ro
  data_files:
  - split: train
    path: ro/train/data-*
  - split: validation
    path: ro/validation/data-*
- config_name: cat
  data_files:
  - split: train
    path: cat/train/data-*
  - split: validation
    path: cat/validation/data-*
- config_name: combined
  data_files:
  - split: train
    path: combined/train/data-*
  - split: validation
    path: combined/validation/data-*
---


# LivingNER: Named entity recognition, normalization & classification of species, pathogens and food


### Dataset Summary

The LivingNER Gold Standard corpus is a collection of 2000 clinical case reports covering a broad range of medical specialities, i.e. infectious diseases (including Covid-19 cases), cardiology, neurology, oncology, dentistry, pediatrics, endocrinology, primary care, allergology, radiology, psychiatry, ophthalmology, urology, internal medicine, emergency and intensive care medicine, tropical medicine, and dermatology annotated with species [SPECIES] (including living organisms and microorganisms) and infectious diseases [ENFERMEDAD] mentions. Species mentions include many pathogens and infectious agents, but also food, allergens, pets or other species, taxonomic groups and organisms of clinical relevance. 

The  LivingNER corpus has also annotations of mentions of humans (tag HUMAN), including the patients itself, family members, healhcare professionals or other persons mentioned in the case reports. Thus it can be useful to extract family history information of patients or information about the social and healthcare personal environment and interactions.

All mentions have been exhaustively manually mapped by experts to their corresponding (NCBI Taxonomy)[https://www.ncbi.nlm.nih.gov/taxonomy] identifiers. 

It was used for the (LivingNER)[https://temu.bsc.es/livingner/] Shared Task on pathogens and living beings detection and normalization in Spanish medical documents, which was celebrated as part of IberLEF 2022.


## Dataset Description

- **Languages:**
  - en
  - fr
  - gal
  - it
  - pt
  - ro
  - combined
- **Training Set Size:** 1000
- **Test Set Size:** 500
- **Features:**
  - text: Original text
  - language: Language identifier
  - tokens: Tokenized text
  - ner_tags: Named entity tags in BIO format
  - entity_mentions: Detailed entity information

## Usage

```python
from datasets import load_dataset

# Load the dataset
dataset = load_dataset('path/to/dataset', '{lang}')

# Access splits
train_data = dataset['train']
test_data = dataset['test']
```

## Labels

The following entity types are annotated in this dataset:
['O', 'B-HUMAN', 'I-HUMAN', 'B-SPECIES', 'I-SPECIES']

### Citation Information
```json
@article{amiranda2022nlp,
  title={Mention detection, normalization \& classification of species, pathogens, humans and food in clinical documents: Overview of LivingNER shared task and resources},
  author={Miranda-Escalada, Antonio and Farr{\'e}-Maduell, Eul{`a}lia and Lima-L{\'o}pez, Salvador and Estrada, Darryl and Gasc{\'o}, Luis and Krallinger, Martin},
  journal = {Procesamiento del Lenguaje Natural}, year={2022}
}

@dataset{miranda_escalada_2022_7684093,
  author       = {Miranda-Escalada, Antonio and
                  Farré-Maduell, Eulàlia and
                  Lima-López, Salvador and
                  González Gacio, Gloria and
                  Krallinger, Martin},
  title        = {LivingNER corpus: Named entity recognition,
                   normalization \& classification of species,
                   pathogens and food
                  },
  month        = jun,
  year         = 2022,
  publisher    = {Zenodo},
  version      = {6.3.1},
  doi          = {10.5281/zenodo.7684093},
  url          = {https://doi.org/10.5281/zenodo.7684093},
}
```