Counterfactual data generation and fine-tuning improve group fairness metrics for Dutch hate speech detection while slightly hurting counterfactual token fairness on toxic content.
In Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society, pages 219–226
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Towards Fairness Assessment of Dutch Hate Speech Detection
Counterfactual data generation and fine-tuning improve group fairness metrics for Dutch hate speech detection while slightly hurting counterfactual token fairness on toxic content.