Politeness
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/tag-and-generate/politeness-dataset
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是为了对文本进行礼貌与否的分类而设计，其来源于恩隆电子邮件语料库。该数据集是一个非平行语料库，规模包括20万个非平行语料，这些语料要么是礼貌风格，要么是不礼貌风格，同时还包括1,000个测试句子。其任务是进行文本分类。
This dataset is developed for the task of classifying texts into polite and impolite categories, and it is sourced from the Enron Email Corpus. As a non-parallel corpus, it consists of 200,000 non-parallel text samples that are either in polite or impolite styles, together with 1,000 test sentences. The core task of this dataset is text classification.
提供机构：
Madaan et al.
