Controversies over the effectiveness of masks, disinformation about supposed miracle cures, various and varied conspiracy theories … Since the beginning of the Covid-19 pandemic, there has been an enormous amount of fake news and poisoning. So much so that large social networks sometimes struggle to clean up. According to a report by the NGO Avaaz, only 16% of fake health messages on Facebook are flagged as such. Twitter, for its part, tries to contain this false information.
Big data and AI, a magic bullet against false news?
The World Health Organization is aware of this phenomenon and has decided to deal with it in its own way. She works with a company to analyze conversations about the coronavirus and uncover misinformation. This social listening strategy consists of digitizing around 1.6 million social media posts every week. Once collected, this data is sifted through a machine learning algorithm to classify the information into four categories: cause, disease, interventions, and treatments.
By identifying topics that are becoming increasingly popular with the public, WHO hopes to hit the mark in its communications and bring its version of the facts to the public debate. The institution is aware of the limitations of its approach as not everyone is present on social networks. She therefore initiated the same social listening program for radio stations. An initiative of this type was attempted in Uganda, for example, to counter rumors that the coronavirus could be treated entirely with medicinal plants.
This is not the first time artificial intelligence and big data have been used to combat fake news. For example, we told you about the campaign for public goods projects. The public health nonprofit has recruited an army of volunteers to engage in online debates about anti-vaccines. The latter benefit from efficient analysis tools for social networks to achieve this goal.