Natural Language Inference with Self-Attention for Veracity Assessment of Pandemic Claims
Natural Language Inference with Self-Attention for Veracity Assessment of Pandemic Claims
复制标题
DOI：
10.48550/arxiv.2205.02596
复制
发表时间：
2022-05
期刊：
影响因子：
--
通讯作者：
M. Arana-Catania;E. Kochkina;A. Zubiaga;M. Liakata;R. Procter;Yulan He
中科院分区：
文献类型：
--
作者：
M. Arana-Catania;E. Kochkina;A. Zubiaga;M. Liakata;R. Procter;Yulan He
我们提出了一项关于自动准确性评估的全面工作，从数据集创建到基于自然语言推理（NLI）开发新方法，重点关注与COVID-19大流行相关的错误信息。我们首先描述了新PANACEA数据集的构建，该数据集由COVID-19的异质声明及其各自的信息源组成。数据集的构建包括检索技术和相似性测量方面的工作，以确保一套独特的索赔。然后，我们提出了基于自然语言推理的自动准确性评估的新技术，包括图卷积网络和基于注意力的方法。我们已经进行了实验的证据检索和准确性评估的数据集上使用所提出的技术，并发现他们的竞争SOTA方法，并提供了详细的讨论。
We present a comprehensive work on automated veracity assessment from dataset creation to developing novel methods based on Natural Language Inference (NLI), focusing on misinformation related to the COVID-19 pandemic. We first describe the construction of the novel PANACEA dataset consisting of heterogeneous claims on COVID-19 and their respective information sources. The dataset construction includes work on retrieval techniques and similarity measurements to ensure a unique set of claims. We then propose novel techniques for automated veracity assessment based on Natural Language Inference including graph convolutional networks and attention based approaches. We have carried out experiments on evidence retrieval and veracity assessment on the dataset using the proposed techniques and found them competitive with SOTA methods, and provided a detailed discussion.