Implicit Abuse Dataset
收藏官方服务：
资源简介：
Implicit Abuse Dataset 1,938 tweets annotated for implicit and explicit abusive language in online celebrity harassment. Labels:- explicit (665): Direct, overt abusive language- implicit (663): Subtle, indirect abuse (sarcasm, coded language)- None (610): Non-abusive tweets Annotation: 3 annotators, Fleiss' kappa = 0.62 Format: UTF-8 CSV with columns: text, label
提供机构：
Zenodo创建时间：
2026-03-11
