Social Media Implicit Hate Speech Dataset
收藏DataCite Commons2026-01-14 更新2026-05-05 收录
下载链接：
https://www.scidb.cn/detail?dataSetId=28c5672acb0f4db7abc34a4fa7d9d956
下载链接
链接失效反馈官方服务：
资源简介：
This dataset is specifically designed to evaluate models' ability to detect disguised hate speech. It is derived from the TOXICLOAKCN dataset, which itself is constructed by applying perturbations—such as homophone substitution, emoji replacement, and Pinyin augmentation—to original sentences from datasets like TOXICN. These perturbations aim to simulate common strategies used on social media to evade content moderation. In this work, we randomly sampled 1,000 instances from this dataset to assess the generalization capability of large language models in identifying implicit hate speech.
提供机构：
Science Data Bank
创建时间：
2026-01-14
