Datasets:
dataset_info:
features:
- name: tokens
sequence: string
- name: ner_tags
sequence:
class_label:
names:
'0': O
'1': B-PESSOA
'2': I-PESSOA
'3': B-LOCAL
'4': I-LOCAL
'5': B-DATA
'6': I-DATA
'7': B-ORGANIZACAO
'8': I-ORGANIZACAO
splits:
- name: train
num_bytes: 2234513
num_examples: 6175
- name: eval
num_bytes: 553739
num_examples: 1544
- name: test
num_bytes: 715658
num_examples: 1930
download_size: 829002
dataset_size: 3503910
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*
- split: test
path: data/test-*
task_categories:
- token-classification
language:
- pt
pretty_name: MariNER
size_categories:
- 1K<n<10K
This is the MariNER (Mapeamento e Anotações de Registros hIstóricos para NER) dataset, a Named Entity Recognition dataset for Brazilian Portuguese in the historical domain, presented in the paper MariNER: A Dataset for Historical Brazilian Portuguese Named Entity Recognition. The dataset consists of five documents from early 20th Century on expeditions through Brazil's Northeast region. The documents were manually annotated with four entity types: Person, Location, Date and Organization. They were also divided into sentences and tokenized.
The dataset is available in the BIO format, with the following distinct labels and corresponding indices:
| Label | Tag |
|---|---|
| 0 | O |
| 1 | B-PESSOA |
| 2 | I-PESSOA |
| 3 | B-LOCAL |
| 4 | I-LOCAL |
| 5 | B-DATA |
| 6 | I-DATA |
| 7 | B-ORGANIZACAO |
| 8 | I-ORGANIZACAO |
The dataset is also available in Github at https://github.com/Joao-Luz/mariner
To cite this work, use:
@misc {
sarcinelli2025marinerdatasethistoricalbrazilian,
title={MariNER: A Dataset for Historical Brazilian Portuguese Named Entity Recognition},
author={João Lucas Luz Lima Sarcinelli and Marina Lages Gonçalves Teixeira and Jade Bortot de Paiva and Diego Furtado Silva},
year={2025},
eprint={2506.23051},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2506.23051},
}