Hatebase - Global network building and artificial intelligence for multilingual hate speech monitoring

Hate speech is widely acknowledged as a precipitator of violence, particularly against vulnerable minorities. It contributes to societal polarization that can lead to instability and conflict. The impact of hate speech on fragile states has grown in recent years because of increased internet connectivity. This enables hate speech to be spread via social media, which contributes to persecution, armed conflict, and genocide in various countries. The sheer volume of hate speech circulating online exceeds the capabilities of human moderators, resulting in the need for increasingly effective automation. The pervasiveness of online hate speech also presents an opportunity, since these large volumes of data are potentially useful as indicators of spiraling instability and they can introduce the possibility of early intervention.

The Sentinel Project, a Canadian NGO that works to prevent genocide using an early-warning system, is implementing this project. The goal is to address online hate speech by focusing on technology, partnerships, and research. In terms of technology, the project will enhance the Hatebase database and monitoring platform, which incorporates elements of artificial intelligence and natural language processing to expand the breadth and capabilities of the algorithm that powers it. Developed by the Sentinel Project, the award-winning application is currently the world's largest structured repository of regionalized, multilingual hate speech.

Partnership efforts will focus on building a global network of research nodes to assist with ongoing multilingual and contextual training on the platform. Research efforts will examine correlations between online hate speech and offline incidents, particularly in fragile contexts, and it will analyze how these can be used to inform mitigation responses to hate speech.