FakeClaim

FakeClaim数据集是由杜伊斯堡-埃森大学等机构创建，专注于2023年以色列-哈马斯战争期间的事实核查。该数据集包含1499条来自60个事实核查组织的多语言事实核查文章，旨在通过自动分类假YouTube视频来识别假新闻。数据集的创建涉及从社交媒体和新闻源中提取声明，并通过训练有素的新闻记者进行验证。FakeClaim数据集的应用领域包括分析战争期间的信息空间，以及改进支持事实核查的技术。

The FakeClaim dataset was developed by institutions including the University of Duisburg-Essen, with a core focus on fact-checking during the 2023 Israel-Hamas war. It comprises 1499 multilingual fact-checking articles sourced from 60 fact-checking organizations, and its primary objective is to identify fake news through automatic classification of deceptive YouTube videos. The creation of the FakeClaim dataset entails extracting claims from social media platforms and news feeds, followed by manual verification conducted by trained professional journalists. Potential application scenarios of the FakeClaim dataset include analyzing the information ecosystem during the 2023 Israel-Hamas conflict, as well as enhancing technologies that support fact-checking work.

搜集汇总

背景与挑战

背景概述

FakeClaim数据集专注于2023年以色列-哈马斯战争期间的事实核查，包含1499条来自60个组织的多语言事实核查文章，旨在自动分类假YouTube视频以识别假新闻。该数据集通过从社交媒体和新闻源提取声明并由记者验证而创建，应用于分析战争信息空间和改进事实核查技术。