GermEval 2021 Training Dataset

该数据集包含了经过标注的Facebook评论，这些评论根据其毒性、互动性和陈述事实的特点进行了分类。评论由经过培训的标注员进行标注，测试数据来源于不同主题的讨论。数据集的规模包括3,244条训练评论和944条测试评论，其任务是对评论进行二分类，将其归入毒性、吸引互动和陈述事实的类别中。

This dataset contains annotated Facebook comments, which are categorized based on three key characteristics: toxicity, interactivity, and fact-stating nature. All comments were annotated by trained human annotators, and the test data is sourced from discussions covering a wide range of topics. The dataset comprises 3,244 training comments and 944 test comments. The primary task associated with this dataset is to perform binary classification on the comments, assigning them to the categories of toxic, interaction-attracting, and fact-stating.