nishant773911/HateXplain
收藏Hugging Face2025-12-11 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/nishant773911/HateXplain
下载链接
链接失效反馈官方服务：
资源简介：
HateXplain是一个用于仇恨言论和冒犯性语言检测的基准数据集，特别之处在于它包含了解释和理由的注释，旨在支持在线内容审核中可解释模型的发展。数据集包含约20,000条英语社交媒体帖子，每条帖子都有多个注释者的标签和理由注释，标签分为normal、offensive或hatespeech。数据集结构包括帖子ID、分词后的帖子内容、注释者信息和理由标记。
HateXplain is a benchmark dataset for hate speech and offensive language detection, uniquely annotated with explanations and rationales. It is designed to support the development of interpretable models in online content moderation. The dataset contains approximately 20,000 English social media posts, each with multiple annotators labels and rationale annotations, categorized as normal, offensive, or hatespeech. The dataset structure includes post ID, tokenized post content, annotator information, and rationale markers.
提供机构：
nishant773911
