HatEval

该数据集名为HatEval，专注于检测针对移民和女性的仇恨言论，它是一个二元分类问题，旨在识别推文中是否包含仇恨言论。原始数据集被划分为训练集、开发集和测试集，作为比赛的一部分。数据集的类别分布均衡，其中仇恨言论占42%，非仇恨言论占58%，其任务是进行仇恨言论的检测。

This dataset, named HatEval, focuses on detecting hate speech targeting immigrants and women. It is a binary classification task aimed at identifying whether a given tweet contains hate speech. The original dataset was split into training, development, and test sets as part of a shared task. The dataset has a relatively balanced class distribution, with hate speech accounting for 42% of the total samples and non-hate speech making up 58%. The core task of this dataset is hate speech detection.