Bridging the gap in online hate speech detection: a comparative analysis of bert and traditional models for homophobic content identification on x/twitter

HIGHLIGHTS

• What: By releasing the largest open-source labelled English dataset for homophobia detection known to the authors an analysis of various models` performance and the strongest BERT-based model the authors aim to enhance online safety and inclusivity. Given the short nature of tweets and that BERT has been found to be effective in sentence-level tasks , the authors attempt to exploit BERT`s effectiveness at sentence-level classification for the homophobia detection model. In the experiments , the authors investigate the impact of BERT embeddings as a feature for classifying homophobic content with traditional models and BERT . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or Email

  Password

  Remember me

  Lost your password?

  If you don't have an account, you can create one here.