Repository for the CREHate dataset, presented in the paper "Exploring Cross-cultural Differences in English Hate Speech Annotations: From Dataset Construction to Analysis". (NAACL 2024)
Most hate speech datasets neglect the cultural diversity within a single language, resulting in a critical shortcoming in hate speech detection.
To address this, we introduce CREHate, a CRoss-cultural English Hate speech dataset.
To construct… See the full description on the dataset page:
https://huggingface.co/datasets/nayeon212/CREHate.
