Affect in Tweets
收藏arXiv2025-09-30 收录
下载链接：
https://competitions.codalab.org/competitions/17751
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了来自推特的数据，其中每条推文都根据推文发布者的11种情绪状态之一或多种进行了分类。具体来说，训练集、验证集和测试集分别包含了6,838个、886个和3,259个样本。整个数据集规模为10,983条推文，所面临的任务是进行细粒度的情绪识别。
This dataset contains data from Twitter, where each tweet is classified based on one or more of the 11 emotional states of its publisher. Specifically, the training, validation, and test sets include 6,838, 886, and 3,259 samples respectively. The entire dataset consists of 10,983 tweets in total, and the underlying task is fine-grained emotion recognition.
提供机构：
SemEval-2018
