Prevalence of Hate Speech on Twitter in Switzerland and Strategies to Counter It

The new project financed by BAKOM combines natural language processing (NLP) and machine learning with civil society engagement to analyse the prevalence of online hate speech on Twitter in Switzerland and to identify effective strategies to counter it. The project is a collaboration between alliance F (Federation of Swiss Women’s Associations), the Digital Democracy Lab (UZH) and the Immigration Policy Lab (ETH).

This project seeks first to carefully analyse the prevalence of hate speech and then to test the effectiveness of counter speech strategies on Twitter in Switzerland. With the help of the hate speech classification algorithm developed as part of the Stop Hate Speech project, the first part of the project focuses on collecting and examining a large corpus of Twitter comments to document the presence and distribution of hate speech on Twitter in Switzerland. In the second part of the project, a randomised field experiment will be used to test the effectiveness of a set of counter speech strategies for reducing hate speech on Twitter.

The close collaboration with alliance F ensures that the scientific results directly contribute to the effective detection and reduction of online hate speech in Switzerland. The goal is to improve the quality of public discourse without resorting to censorship and to minimise offline consequences of hostile online behaviour.

Involved in the project from the side of the DSI Democracy Community are Prof. Fabrizio GilardiProf. Karsten Donnay and Dr. Ana Kotarcic.