The threat from online extremism is growing. In the wake of Charlottesville and other incidents, the public outcry over extremist recruitment and messaging online has reached a tipping point. As extreme ideologues become ever more savvy at mobilizing via social media, the calls for mitigating hateful activity on the social networks will grow ever more intense. Yet these discussions should be based on data and insight rather than fear and intuition.
With this in mind, we designed an interactive dashboard to monitor and study extremist activity on Twitter. The dashboard focuses on just over 1,000 accounts that regularly use hateful content against protected groups. Our approach provides a visual and qualitative understanding of the scale and scope of online hateful activity in near real time.
The dashboard makes it possible to better understand whether extremists are trying to shape a narrative or merely responding to world events. Therefore, it allows for us to study how they drive engagement. Do they use an echo chamber approach, in which extremist accounts reply and interact with each other exclusively? Or do they primarily seek external engagement instead – for example, by hijacking popular hashtags and piggybacking on top of other networks?
The value of this dashboard is not only that it can provide snapshots of hateful activity online in near real time, but also that it can provide a historical benchmark. When does extremist activity peak? And can we reliably predict that protests or events are imminent?