LIARx is a partial fact dataset for fake news detection in the domain of US politics and is an updated form of the popular LIAR dataset published by William Yang Wang in 2017. The dataset addresses the non-binary nature of fake news and instead allows you to learn the degree of veracity of each instance. README.md (markdown format) provides further information about the dataset and its features.

{"references": ["Wang, W.Y.: \" liar, liar pants on fire\": A new benchmark dataset for fake newsdetection. arXiv preprint arXiv:1705.00648 (2017)"]}