Staying alone together: developing fake news immunity - from fallacies to fake news

21^st June 2021, 16:00
Elena Musi

Abstract

One of the major challenges of the current information ecosystem is the rapid spread of misinformation through digital media. Even though unintentionally dangerous, misinformation has a wide societal impact: 59% of fake news do not contain neither fabricated nor imposter content, but rather reconfigured misinformation (Brennen et al., 2020), which proliferates through social media, the main source of news for infodemically vulnerable citizens. However, the identification of misinformation is far from being successfully addressed by human fact-checkers, let alone automatic ones due to the lack of a common truth barometer that hinders the creation of datasets to train automatic systems. We believe that what makes these types of news ‘fake’ is not the mere truth of the information conveyed, but the fallacious way of presenting the arguments they contain through false analogies, hasty generalizations, and cherry picking of information (Musi and Reed, under review).

In our UKRI funded project Being Alone Together: Developing Fake News Immunity (https://fakenewsimmunity.liverpool.ac.uk/), we propose to counter misinformation providing citizens with the means to act as their own fact-checkers to avoid creating and spreading misleading news. Drawing from the multi-level annotation of a dataset of 1500 COVID-19 related news web-crawled from 5 English fact-checkers, we propose a systematic procedure to identify fallacious arguments across different digital media sources and type of claims (e.g. predictions, interpretations). Leveraging the outcomes of data analysis we built two chatbots that we would like to present at the symposium, the Fake News Immunity Chatbot (http://fni.arg.tech/) and the Vaccinating News Chatbot (http://fni.arg.tech/?chatbot_type=vaccine), respectively targeting citizens and communication gatekeepers.