RumorEval

该数据集名为RumorEval，包含了八个突发新闻事件，专为SemEval-2017任务8竞赛而开发。每个主张都被标记为二元类别（真谣言或假谣言）。由于该数据集是为竞赛而开发的，因此它对于谣言检测研究的相关性得到了增强。数据集的规模涵盖了多个事件（共八个突发新闻事件），其任务专注于谣言检测。

This dataset is named RumorEval. It includes eight breaking news events and was specifically developed for the SemEval-2017 Task 8 competition. Each claim is annotated with a binary classification label: genuine rumor or fake rumor. Since this dataset was developed for a competition, its relevance to rumor detection research has been significantly enhanced. Covering a total of eight breaking news events, the dataset is focused on rumor detection tasks.

搜集汇总

背景与挑战

背景概述

RumorEval是SemEval-2017 Task 8的数据集，专注于确定谣言的真实性并提供支持证据，用于社交媒体谣言检测研究。数据集基于PHEME谣言方案，包含训练、开发和测试数据，涉及两个子任务：任务A（谣言验证）和任务B（谣言支持分析），数据来源于推特对话线程和新闻故事。