Hatespeech-Dataset / README.md
nikesh66's picture
Create README.md
4349651
|
raw
history blame
466 Bytes
metadata
language:
  - en

Hate Speech Dataset

This dataset contains artificially genrated tweets alongwith its label whether it is hatespeech or not

Dataset Description

  • Number of Rows: 5,000
  • Number of Columns: 2
  • Column Names: 'Tweet', 'Hate Speech'
  • Description: This dataset comprises tweets with annotations indicating whether they contain hate speech or not. Each row has a tweet and a binary label ('yes' or 'no') denoting the presence of hate speech.