2026-10-11T21:09:27Z https://www.scidb.cn/oai

10.57760/sciencedb.j00133.00632 2026-01-14T08:48:17Z

2026-01-14 Social Media Implicit Hate Speech Dataset doi:10.57760/sciencedb.j00133.00632 en 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. social media; Hate Speech Detection; Large Language Models yuan wei kang RESTRICTED dataset Science Data Bank