The World Health Organization (WHO) acknowledged that: “The 2019-nCoV outbreak and response has been accompanied by a massive ‘infodemic’ … that makes it harder for people to find trustworthy sources and reliable guidance when they need it”.
In this lecture, speakers will discuss tools developed by QCRI to help address misinformation and information manipulation in general in the context of social media and traditional media outlets.
Using Twitter as an example, speakers will discuss the politicization of the COVID-19 pandemic, while demonstrating QCRI-developed tools that automatically identify the political leanings of users and approximate the polarization of parties’ supporters based on the content of their tweets. Others map the language of different groups along multiple linguistic dimensions such as sentiment and topicality. This in turn will help investigate how the COVID-19 pandemic is being politicized and the effect on the perception of medical realities.
Speakers will also present the “Tanbih” news aggregator, which aims to limit the effect of "fake news", propaganda, and media bias by making users aware of what they are reading. The project's primary aim is to promote media literacy and critical thinking, which are arguably the best ways to address misinformation in the long run. Through this platform, one can develop media profiles showcasing factuality of reporting, the degree of propagandistic content, hyper-partisanship, leading political ideology, general frame of reporting, stance with respect to various claims and topics, as well as audience reach and audience bias in social media. Speakers will further demonstrate the automatic detection of specific propaganda techniques being used in the news and how Tanbih’s tools can help in the fight against COVID-19 misinformation.
The lecture will conclude with a call for action and our latest research on performing fine-grained analysis of COVID-19 social media posts that combine the perspectives and the interests of journalists, fact-checkers, social media platforms, policy makers, and society as a whole.
Kareem Darwish is a principal scientist and the acting managing director of the Arabic Language Technologies at QCRI with an interest in natural language processing (NLP), social computing, and information retrieval. He is currently developing a state-of-the-art Arabic NLP toolkit, which includes POS tagging, named entity recognition, parsing, etc. Darwish is also working on the automated detection of propaganda accounts on social media and stance detection in social computing.
Preslav Nakov is a principal scientist at QCRI. His research interests include computational linguistics and NLP, disinformation, propaganda, fake news and bias detection, fact-checking, machine translation, question answering, sentiment analysis, lexical semantics, web as a corpus, and biomedical text processing.
He is the Principal Investigator (PI) of the QCRI mega-project Tanbih. He is also the lead-PI of a QCRI-MIT collaboration project on Arabic Speech and Language Processing for Cross-Language Information Search and Fact Verification.
Giovanni Da San Martino is a scientist at QCRI, whose research interests are at the intersection of machine learning and NLP, with applications for community question answering and news analysis.
Since 2018, he has been involved in the Tanbih and Catalyst projects, and in building intelligent tools for news analysis in collaboration with MIT-CSAIL and media partners such as Al Jazeera and Associated Press.
This lecture will be held in English.