Skip to content

Dataset · peluz

Legal named entity recognition dataset

peluz/lener_br

Manually annotated Brazilian laws and court cases for named entity recognition in Portuguese.

What's inside

The corpus consists of 70 official Portuguese documents, comprising 66 judicial acts from Brazilian courts and 4 legislative files. Texts are manually annotated with tags for persons, locations, time periods, organizations, laws, and court cases.

What you can do with it

  • Training named entity recognition (NER) models on legal texts
  • Automated indexing and retrieval of references to laws and court cases
Tasks
Token classification
Languages
Portuguese
Size
10K–100K rows
Downloads this month
1.2K
Likes
42

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("peluz/lener_br")