RumourEval 2019 data

Manual stance and veracity judgments over a large set of English social media conversations. Judgments are expertsourced and crowdsourced with extensive quality control as detailed in the referenced paper. Conversations are around a central claim; claims are grouped into news themes, each theme centering on a current event. Social media platforms include Twitter and Reddit. This is the dataset for the SemEval-2019 task, RumourEval.

针对大规模英文社交媒体对话集合开展人工立场判定与真实性核验。该标注工作由专家与众包人员共同完成，并辅以严格的质量管控流程，相关细节详见所引用的论文。所有对话均围绕某一核心主张展开；这些主张被归类至不同新闻主题，每个主题均聚焦于一则时事事件。涉及的社交媒体平台包括推特（Twitter）与红迪（Reddit）。本数据集即为SemEval-2019任务RumourEval所使用的数据集。