Viral hate: Who is exposed to hatred?

From Survey to artificial intelligence in the study of online political debates, Tobias Bornakke, Analyse & Tal

Surveys are often the foundation for how we understand society and societal changes. Where the survey method excels in its ability to capture our reflections upon a given subject, it often comes short when we try to capture and describe behavior over time. In this presentation we show how artificial intelligence, in the shape of a machine learning algorithm (in Danish) to identification of hateful language, offers a new and different view on the tone in the online political debate. The algorithm enables us to follow what political issues that especially provokes hate speech while also offering a more nuanced insight into the extent and development of hate speech over time. We conclude the presentation by discussing how studies of behavior enables us to raise new societal agendas and perspectives.