Explainability and Hate Speech

Official repository for "Explainability and Hate Speech: Structured Explanations Make Social Media Moderators Faster".

Annotation Scheme

The file annotations.csv contains the collected annotations after filtering out the outliers (see papers for details).

The csv file contains the following columns:

• annotator: the numerical ID assigned to the moderator who completed the annotation.
• instance: the ID of the annotated post. Post IDs are taken from the PLEAD dataset.
• length: the number of characters in the post.
• scenario: the ID of the scenario. s1 corresponds to the scenario where no explanations were shown, s2 corresponds to the scenario with generic explanations and s3 refers to the scenario with structured explanations.
• time: the annotation time in seconds.
• label: 1 for posts that violate the policy, 0 for posts that comply with the policy.
• prediction: 1 if the moderator thought the post violates the policy, 0 otherwise.

Reference

If you use our dataset or model, please cite our paper:

@article{calabrese-etal-2024-structured-explanations,
  author    = {Agostina Calabrese and
               Leonardo Neves and
               Neil Shah and
               Maarten W. Bos and
               Bj{\"{o}}rn Ross and
               Mirella Lapata and
               Francesco Barbieri},
  title     = {Explainability and Hate Speech: Structured Explanations Make Social Media Moderators Faster},
  journal   = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics},
  year      = {2024}
}