A European Observatory for Social Media Content

Guido Caldarelli, Professor in Theoretical Physics, Sapienza University of Rome

Fake news, alternative facts, post-truth era, circular reporting, echo chambers, have nowadays become very popular in journals and TV news. The number of headlines that pop-up each day on fake news and the challenges of today’s media environment show how timely this issue is and how pressing it is to come with a systematic, empirical and theoretical approach to deal with the issue of how (mis)information spreads on social media.

Therefore, studying the mechanisms that shape social media and their impact on society will allow to respond to the key challenges posed by the increasingly broad and diverse community of stakeholders interested in this topic. These include researchers, policymakers, regulators, journalists, media companies (traditional and new), industry, civil society organisations and citizens.

Starting from an evaluation of the state of the art of current and ongoing literature and research on this topic (and keeping in mind that, today, no single solution satisfactorily addresses the issues implied by the rapid diffusion of content on social media), it is necessary to build a pilot European Observatory that can gather and analyse content from different social media. This would make it possible, to a certain extent, to identify, select and differentiate true from false content, to learn how this content is produced and distributed, and how citizen’s react to it (which narratives are more trusted, why, by whom…).

The Observatory should also provide tools to visualise and analyse the dynamics of news spreading; experiment algorithms for defining an online reputation for news producers; use human behaviour and null models to test the possibility to forecast outcomes, and compare with evidence in data.

Micro-targeting should also be investigated in its current practices, with a view to, on the one hand, showing how personality and behaviours can be predicted on the basis of online data and, on the other hand, raising users’ awareness about the existence of these practices.

The issue is particularly sensitive, since we do not want to create a ‘Ministry of truth’. We should push for an ecosystem of different fact checkers that can help the debunking procedure. We should rapidly prepare a list of questions to target (i.e. one or more app to check if the twitter user is a bot or not) and ask practitioners and scientists to develop the instruments necessary.

Interesting further readings include: http://arxiv. org/abs/1603.01511; http://cnets.indiana.edu/ blog/2016/12/21/hoaxy/; https://politoscope. org/2018/09/publication-inaugurale-politoscope