Keeping up-to-date on the topics spreading through social media for tip-generation and fact-checking can quickly become unmanageable given the size and scale of the data. For closed-messaging platforms such as WhatsApp, where 40 billion messages are sent per day, it is nearly impossible without a combination of automation with human intelligence. This panel will discuss how automated ingestion and duplicate detection of content from platforms can be used to create monitoring dashboards that enable fact-checkers and journalists to quickly see what content, topics, etc. are getting attention on these platforms.
Will (Full Fact) is leading automated fact-checking and researching how the impact/spread of statements by politicians on social media and the traditional media can be extracted and summarised to help fact-checkers decide which claims to investigate. Kiran (MIT) built a “WhatsApp monitor” displaying the most shared content around the Brazilian and Indian elections from WhatsApp groups that published join links publically on the Internet that was actively used by journalists reporting on these events. Scott (Meedan) oversees research into datasets created through running WhatsApp tiplines where people can submit content to teams of fact-checkers.
Organised in association with Meedan.