---
license: cc-by-nc-sa-4.0
task_categories:
- text-classification
- zero-shot-classification
language:
- en
tags:
- hate speech
- abuse detection
- toxicity
- robustness
- AAA
size_categories:
- 1K<n<10K
---

# PLEAD-AAA
This test set was generated applying the [AAA tool](https://github.com/Ago3/Adversifier) to the [PLEAD](https://huggingface.co/datasets/agostina3/PLEAD) dataset.

## Reference
If you use our dataset, please cite our papers:
```
@article{calabrese-etal-2022-plead,
  author    = {Agostina Calabrese and
               Bj{\"{o}}rn Ross and
               Mirella Lapata},
  title     = {Explainable Abuse Detection as Intent Classification and Slot Filling},
  journal   = {Transactions of the Association for Computational Linguistics},
  year      = {2022}
}

@inproceedings{calabrese2021aaa,
  title={Aaa: Fair evaluation for abuse detection systems wanted},
  author={Calabrese, Agostina and Bevilacqua, Michele and Ross, Bj{\"o}rn and Tripodi, Rocco and Navigli, Roberto},
  booktitle={Proceedings of the 13th ACM Web Science Conference 2021},
  pages={243--252},
  year={2021}
}
```