Cross-SEAN: A Cross-Stitch Semi-Supervised Neural Attention Model for COVID-19 Fake News Detection

As the COVID-19 pandemic sweeps across the world, it has been accompanied by\na tsunami of fake news and misinformation on social media. At the time when\nreliable information is vital for public health and safety, COVID-19 related\nfake news has been spreading even faster than the facts. During times such as\nthe COVID-19 pandemic, fake news can not only cause intellectual confusion but\ncan also place lives of people at risk. This calls for an immediate need to\ncontain the spread of such misinformation on social media. We introduce CTF,\nthe first COVID-19 Twitter fake news dataset with labeled genuine and fake\ntweets. Additionally, we propose Cross-SEAN, a cross-stitch based\nsemi-supervised end-to-end neural attention model, which leverages the large\namount of unlabelled data. Cross-SEAN partially generalises to emerging fake\nnews as it learns from relevant external knowledge. We compare Cross-SEAN with\nseven state-of-the-art fake news detection methods. We observe that it achieves\n$0.95$ F1 Score on CTF, outperforming the best baseline by $9\\%$. We also\ndevelop Chrome-SEAN, a Cross-SEAN based chrome extension for real-time\ndetection of fake tweets.\n