SemEval-2023 Task 3 Datasets
收藏arXiv2025-09-30 收录
下载链接：
https://competitions.codalab.org/competitions/20235#learn_the_details-datasets
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含三个子任务，主要关注跨多种语言的新闻文章分类、框架检测以及说服技巧检测。具体来说，该数据集为每种语言和子任务提供了训练、开发和测试的子集，并包含了数据增强和多语言数据集构建的技术支持。该数据集的规模覆盖了子任务和语言在内的27种不同的测试设置。任务的总体目标是进行新闻体裁、框架以及说服技巧的分类和检测。
This dataset comprises three subtasks, focusing primarily on news article classification, frame detection, and persuasion technique detection across multiple languages. Specifically, this dataset provides training, development, and test subsets for each language and subtask, and includes technical support for data augmentation and multilingual dataset construction. This dataset covers 27 distinct test configurations spanning subtasks and languages. The overall goal of these tasks is to classify and detect news genres, frames, and persuasion techniques.
提供机构：
SemEval organizers
