Didier HENRY (LAMIA) : Predictive models for early detection of Twitter hoaxes

Nowadays, social media are widely used on a daily basis to access news. Indeed, social media networks allow for a rapid and wide dissemination of news. Unfortunately, these platforms used by millions of people are not immune to misinformation, as anyone can be a source of information and there are no trust indicators for users. Rumours of celebrity deaths on social networks can spread very widely in a short time and are difficult to verify. Such rumours can lead to worrying or stressful situations especially for the celebrity\'s relatives and can also have economic or political repercussions. Many contributions have focused on the problem of modelling the spread of rumours by limiting themselves to the analogy of the spread process with the epidemiological phenomenon. In contrast to the vast majority of works that focus on the diffusion of rumours by proposing increasingly complex models, we have adopted a predictive modelling approach based on concrete examples in order to detect cases of hoaxes concerning the announcement of the death of celebrities on Twitter quickly after their diffusion has begun.