Davidson Dataset

该数据集是用于评估仇恨言论检测方法的，结果显示，基于嵌入技术的不同方法在性能上存在显著差异。特别是，该数据集在使用基于变换器（transformer）的模型时，展现出了极高的性能，这些模型取得了最佳的成绩。该任务的目标是仇恨言论检测。This dataset is designed for evaluating hate speech detection methods. The results show that different embedding-based methods exhibit significant performance differences. Specifically, transformer-based models achieve exceptionally high performance when tested on this dataset, securing the best overall results. The core objective of this task is hate speech detection.