FRACAS

FRACAS是一个法语标注的引用关系数据集，由格勒诺布尔阿尔卑斯大学创建。该数据集包含1676篇新闻文本中的10965个引用关系，涵盖直接、间接和混合引用类型。数据集的创建过程涉及手动标注，使用BRAT软件进行，旨在解决新闻中引用提取和来源归属的问题。该数据集的应用领域包括社会科学和自然语言处理，特别是在事实核查、假新闻检测和引用传播追踪方面具有重要价值。

FRACAS is a French-labeled citation relation dataset developed by the University of Grenoble Alpes. It encompasses 10,965 citation relations extracted from 1,676 news texts, covering direct, indirect, and mixed citation types. The dataset was constructed via manual annotation using the BRAT software, with the core objective of addressing the challenges of citation extraction and source attribution in news articles. Its applicable fields span social sciences and natural language processing, and it holds significant research value particularly in the areas of fact-checking, fake news detection, and citation propagation tracking.