One of the 11 scientific articles to be presented at the International Joint Conference on Artificial Intelligence (IJCAI), in Montreal, is authored by members of the Técnico community. At this conference, to be held from 16 and 22 August, Filipe Altoe, a PhD student in Computer Science and Engineering – who co-authored the article with Sérgio Pinto, a master’s student in Computer Science and Engineering, and Sofia Pinto, a professor at Técnico and researcher at Instituto de Engenharia de Sistemas e Computadores – Investigação e Desenvolvimento em Lisboa (INESC-ID) – will present an AI model developed to streamline the verification of digital content facts, helping journalists combat the phenomenon of fake news.
The article “Explainable Automatic Fact-Checking for Journalists Augmentation in the Wild” introduces an AI system that automates all the fact-checking steps previously performed manually by professional journalists. The system requires only the identification of the statement to be verified. Users have access to the entire process, including the evidence gathered, the sources used, and a verdict based on this information. Additionally, the model generates an article explaining the reasoning behind the verdict.
“The system was designed to increase the capacity of professional journalists and not to replace them”, says Filipe Altoe, who explains that the work aims to “speed up the fact-checking process”. What once took days and allowed false information to spread virally “can now be completed in just 10 minutes”, adds the PhD student, who believes this is a “significant tool in the fight against fake news, as it can help reduce the duration of viral misinformation”.
The research included a user study with over a hundred professional journalists to validate the model’s functionality and assess the quality of the explanations it generates.
At the IJCAI conference, only 11 papers from around the world were accepted for presentation, which “provided a sense of validation” for the Técnico team behind this article, according to Filipe Altoe. “We believed our work could significantly contribute to addressing the issue of fake news, which we consider one of the most pressing challenges in modern society – validation at this level confirms that belief”, adds the PhD student.
