README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: int64
    - name: tweet_text
      dtype: string
    - name: tweet_date
      dtype: string
    - name: sentiment
      dtype: int64
    - name: query_used
      dtype: string
  splits:
    - name: Train
      num_bytes: 15324413
      num_examples: 95000
    - name: Validation
      num_bytes: 808825
      num_examples: 5000
    - name: TestTema
      num_bytes: 1057904
      num_examples: 5000
    - name: Test
      num_bytes: 684589
      num_examples: 5000
    - name: Test3classes
      num_bytes: 744334
      num_examples: 4999
  download_size: 11554103
  dataset_size: 18620065
configs:
  - config_name: default
    data_files:
      - split: Train
        path: data/Train-*
      - split: Validation
        path: data/Validation-*
      - split: TestTema
        path: data/TestTema-*
      - split: Test
        path: data/Test-*
      - split: Test3classes
        path: data/Test3classes-*
license: cc-by-nc-4.0
task_categories:
  - text-classification
language:
  - pt
tags:
  - Twitter
  - Sentiment Analysis
  - With Bias
  - Emojis
  - URLs
  - Emoticons
pretty_name: Dataset with Tweets in Portuguese Labelled for Sentiment Analysis.
size_categories:
  - 100K<n<1M

This Dataset is saved by Students of the Subject: Interação em Linguagem Natural 2023/24 From: FCUL (Faculdade de Ciências da Universidade de Lisboa)

Used for Sentiment Analysis with a pre-trained Roberta Model.