Data Driven Analysis of Protest Dynamics: Comparing Black Lives Matter Movement 2020 and South African Unrest 2021 on Twitter

Social media has revolutionized public discourse and activism, fostering digital connectivity and engagement. This study explored predictive modelling of protest event signatures on social media, applying machine learning to identify patterns in digital activism discourse. Data from the Black Lives Matter Movement (BLMM 2020) and the South African Unrest (SAU 2021) were analysed to examine temporal dynamics and public engagement on Twitter. Findings from BLMM tweet activity revealed peaks during global solidarity rallies and troughs during police escalations, showcasing significant variability. Similarly, daily tweet volumes during the SAU demonstrated fluctuations in public interest, with spikes linked to pivotal events. The development of a Protest Social Media Archetype further uncovered underlying patterns and similarities across protest movements. This research underscored social media’s role in amplifying marginalized voices, challenging dominant narratives, and mobilizing collective action. By advancing methodological approaches to studying social media data, it provided insights into online discourse during socio-political events. The findings hold implications for academia, policy-making, and activism, advocating for inclusive and participatory digital advocacy strategies.