Twitter users who share news from outlets known to publish false information (fake news) also share some news from reputable sources online, including sources to which they may be ideologically opposed. Why? To study the news sharing behavior of misinformation spreaders on Twitter, we first identify news from reputable sources that are significantly co-shared with fake news. Then we test one hypothesis for why such news might be shared by misinformation spreaders: in order to fit existing narratives that characterize misinformation content. As compared to users who do not share fake news, tweets by misinformation spreaders sharing news from reputable sources make use of misinformation narratives, i.e., such tweets do serve a narrative-fitting function. Apart from showing that one of the reasons why such information share happens is to fit narratives, we look at how mainstream news articles might be used to support specific misinformation narratives. Misinformation is not just a collection of false statements but is often part of bigger narratives and world views. By studying the co-opting of news that might not contain anything false, but can still be used to support narratives present in misinformation content, we can gain a better understanding of the world of online misinformation.
Pranav Goel is a 4th-year Ph.D. student in Computer Science at the University of Maryland, College Park. He is advised by Prof. Philip Resnik as part of the Computational Linguistics and Information Processing (CLIP) Lab, often working with collaborators (especially political and social scientists) from other departments, labs at other universities, and research groups outside of academia. His research lies at the intersection of Natural Language Processing and Computational Social Science, including analysis of language use in sociopolitical contexts focusing on agenda-setting and framing, enabling better computer-assisted content analysis, and evaluating NLP tools and methods in ways that concord with real-world usage. Some of the latest application focuses of his works include US congressional rhetoric and political discussions on social media, online misinformation, and possible tools and implications for journalism.
Pranav Goel is a 4th-year Ph.D. student in Computer Science at the University of Maryland, College Park. He is advised by Prof. Philip Resnik as part of the Computational Linguistics and Information Processing (CLIP) Lab, often working with collaborators (especially political and social scientists) from other departments, labs at other universities, and research groups outside of academia. His research lies at the intersection of Natural Language Processing and Computational Social Science, including analysis of language use in sociopolitical contexts focusing on agenda-setting and framing, enabling better computer-assisted content analysis, and evaluating NLP tools and methods in ways that concord with real-world usage. Some of the latest application focuses of his works include US congressional rhetoric and political discussions on social media, online misinformation, and possible tools and implications for journalism.
