CoolFace
Datasetpublic

aplominski/ucberkeley-measuring-hate-speech-3class

This is modified version of UCBerkley DLab dataset, converted to represtent harmfull level in 3 classes: SAFE - hate_speech_score < -1 BORDERLINE - -1 < hate_speech_score < 0.5 HARMFUL - hate_speech_score >= 0.5 Text and harmful level classification was created by UCBerkley DLab Team! I'm not creator of this data - i'm only converted it into classes!

sourceHugging Facecc-by-4.0updated 3mo agoView on Hugging Face
0likes26downloads
Dataset Card

This is modified version of UCBerkley DLab dataset, converted to represtent harmfull level in 3 classes:

  1. 1.SAFE - hatespeechscore < -1
  2. 2.BORDERLINE - -1 < hatespeechscore < 0.5
  3. 3.HARMFUL - hatespeechscore >= 0.5

Text and harmful level classification was created by UCBerkley DLab Team! I'm not creator of this data - i'm only converted it into classes!