metadata

configs:
  - config_name: fold0
    data_files:
      - split: train
        path: fold0/train-*
      - split: validation
        path: fold0/validation-*
      - split: test
        path: fold0/test-*
  - config_name: fold1
    data_files:
      - split: train
        path: fold1/train-*
      - split: validation
        path: fold1/validation-*
      - split: test
        path: fold1/test-*
  - config_name: fold2
    data_files:
      - split: train
        path: fold2/train-*
      - split: validation
        path: fold2/validation-*
      - split: test
        path: fold2/test-*
  - config_name: fold3
    data_files:
      - split: train
        path: fold3/train-*
      - split: validation
        path: fold3/validation-*
      - split: test
        path: fold3/test-*
  - config_name: fold4
    data_files:
      - split: train
        path: fold4/train-*
      - split: validation
        path: fold4/validation-*
      - split: test
        path: fold4/test-*
  - config_name: fold5
    data_files:
      - split: train
        path: fold5/train-*
      - split: validation
        path: fold5/validation-*
      - split: test
        path: fold5/test-*
  - config_name: fold6
    data_files:
      - split: train
        path: fold6/train-*
      - split: validation
        path: fold6/validation-*
      - split: test
        path: fold6/test-*
  - config_name: fold7
    data_files:
      - split: train
        path: fold7/train-*
      - split: validation
        path: fold7/validation-*
      - split: test
        path: fold7/test-*
  - config_name: fold8
    data_files:
      - split: train
        path: fold8/train-*
      - split: validation
        path: fold8/validation-*
      - split: test
        path: fold8/test-*
  - config_name: fold9
    data_files:
      - split: train
        path: fold9/train-*
      - split: validation
        path: fold9/validation-*
      - split: test
        path: fold9/test-*
license: mit
task_categories:
  - text-classification
language:
  - pt
tags:
  - legal
  - legislative
dataset_info:
  - config_name: fold0
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32604012
        num_examples: 126228
      - name: validation
        num_bytes: 3576665
        num_examples: 13895
      - name: test
        num_bytes: 4029162
        num_examples: 15579
    download_size: 15179616
    dataset_size: 40209839
  - config_name: fold1
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32520988
        num_examples: 126047
      - name: validation
        num_bytes: 3659115
        num_examples: 14049
      - name: test
        num_bytes: 4029736
        num_examples: 15606
    download_size: 15196083
    dataset_size: 40209839
  - config_name: fold2
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32633002
        num_examples: 126129
      - name: validation
        num_bytes: 3585296
        num_examples: 13997
      - name: test
        num_bytes: 3991541
        num_examples: 15576
    download_size: 15186410
    dataset_size: 40209839
  - config_name: fold3
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32627688
        num_examples: 126247
      - name: validation
        num_bytes: 3578106
        num_examples: 13963
      - name: test
        num_bytes: 4004045
        num_examples: 15492
    download_size: 15180222
    dataset_size: 40209839
  - config_name: fold4
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32597950
        num_examples: 126125
      - name: validation
        num_bytes: 3595374
        num_examples: 14008
      - name: test
        num_bytes: 4016515
        num_examples: 15569
    download_size: 15190975
    dataset_size: 40209839
  - config_name: fold5
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32513117
        num_examples: 126046
      - name: validation
        num_bytes: 3638100
        num_examples: 14047
      - name: test
        num_bytes: 4058622
        num_examples: 15609
    download_size: 15194510
    dataset_size: 40209839
  - config_name: fold6
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32593926
        num_examples: 126307
      - name: validation
        num_bytes: 3597549
        num_examples: 13957
      - name: test
        num_bytes: 4018364
        num_examples: 15438
    download_size: 15184065
    dataset_size: 40209839
  - config_name: fold7
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32580938
        num_examples: 126071
      - name: validation
        num_bytes: 3602352
        num_examples: 14009
      - name: test
        num_bytes: 4026549
        num_examples: 15622
    download_size: 15193564
    dataset_size: 40209839
  - config_name: fold8
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32569125
        num_examples: 126146
      - name: validation
        num_bytes: 3612787
        num_examples: 13966
      - name: test
        num_bytes: 4027927
        num_examples: 15590
    download_size: 15172702
    dataset_size: 40209839
  - config_name: fold9
    features:
      - name: id_API
        dtype: string
      - name: summary
        dtype: string
      - name: subject
        sequence: string
    splits:
      - name: train
        num_bytes: 32590719
        num_examples: 126145
      - name: validation
        num_bytes: 3611742
        num_examples: 13936
      - name: test
        num_bytes: 4007378
        num_examples: 15621
    download_size: 15180407
    dataset_size: 40209839

🇧🇷 Brazilian Legislative Bills – Summary Dataset

This dataset contains summaries (ementas) of legislative bills proposed in the Brazilian Chamber of Deputies (BCoD) from 1991 to 2022.
It is intended for multi-label classification, where each bill may be associated with one or more subject categories (temas).

🔀 This is the summary version of the dataset.
If you are looking for the keywords version, see:
👉 ronunes/LegiSubject-Br-Keywords

📁 Dataset Structure

The dataset is organized into 10 stratified cross-validation folds (fold0 to fold9).
Each fold contains 3 standard splits:

train
validation
test

Each split contains the following fields:

Column	Description
`id_API`	Unique identifier for the bill in the BCoD API
`summary`	The summary of the bill (originally `ementa`)
`subject`	List of subject labels associated with the bill

The task is to predict the subject(s) of a bill given its summary.

Usage

You can load each fold easily with the Hugging Face datasets library:

from datasets import load_dataset
ds = load_dataset("ronunes/LegiSubject-Br-Summaries", name="fold0", split="train")