Ax-to-Grind Urdu
收藏arXiv2024-03-21 更新2024-06-21 收录
下载链接：
https://github.com/Sheetal83/Ax-to-Grind-Urdu-Dataset
下载链接
链接失效反馈官方服务：
资源简介：
Ax-to-Grind Urdu是首个大规模公开的乌尔都语假新闻检测数据集，由武汉大学网络空间安全学院创建。该数据集包含10,083条来自巴基斯坦和印度主流乌尔都语报纸和新闻网站的新闻，涵盖15个不同领域，从2017年至2023年。数据集的创建过程包括从网站抓取新闻，并通过专家记者进行手动标注。该数据集主要用于解决乌尔都语假新闻检测问题，旨在通过提供丰富的多领域数据来提高检测算法的准确性和鲁棒性。
Ax-to-Grind Urdu is the first large-scale publicly available Urdu fake news detection dataset, developed by the School of Cyberspace Security, Wuhan University. This dataset comprises 10,083 news articles sourced from mainstream Urdu newspapers and news websites in Pakistan and India, spanning 15 distinct domains and covering the period from 2017 to 2023. The dataset construction process involves scraping news from websites and manual annotation by professional journalists. This dataset is primarily designed to tackle the problem of Urdu fake news detection, aiming to improve the accuracy and robustness of detection algorithms by providing rich multi-domain data.
提供机构：
武汉大学网络空间安全学院
创建时间：
2024-03-21
