Early Warning Signals of Social Instabilities in Twitter Data

The problem concerns the individuation of early warning signals for the outburst of revolutions, riots, or wars. It has been proven in numerous studies - Lim, 2011, 2012; Lotan et al., 2011; Tufekci & Wilson, 2012; Voutaz & Blarer, 2022 - that it is possible to extract early warning signals of contemporary activism from social media. Examples of such events are the Arab Spring of the Tahrir square in 2011 and farmers' protests in India and Iraq of the years 2019 and 2020. Social media such as https://twitter.com/?lang=en are becoming more and more present in everydays life and are often used by people as a mean to express their thoughts or concerns. Hence, tweets likely contain valuable information to identify early warning signals for disruptive events. In this project, we will try to create and study novel techniques to identify early warning signals on tweet data.