Hate Speech

Hate Speech: From automatic classification to understanding emotional dynamics

Hate speech, i.e., verbal aggression targeting members of various social groups, is widespread in everyday life. It has detrimental consequences for everybody exposed to it. Our objective is to understand both, the underlying properties of hate speech, and its impact on listeners.

The existing work on Hate Speech has predominantly been performed on texts such as posts or comments in the social media platforms. Our aim is to study the efficiency of audio recordings in eliciting an emotional response, focusing particularly on the desensitisation in the listener, compared to textual representations. We will create an audio database and analyse acoustic, linguistic, and paralinguistic features relevant to fully automated identification of hate speech. 

The project blends two major methodological approaches. First, experimental studies will manipulate the content of stimuli, their modality, and participants’ response options. Second, signal processing and machine learning methods will be leveraged to identify and differentiate hate speech from other types of speech for both speakers and listeners. Finally, we will study means by which the effects of hate speech on listeners might be reduced through a series of psychological studies. For example, via stopping to listen to it, or by actively generating counter-arguments to hate speech.

We expect that other researchers in the social sciences will be highly interested in using the outcomes of the proposed project for their own work. Conversely, for the speech community in computer science, the manually annotated database will provide a highly useful ground truth for the study and classification of extreme emotions on the Internet.

Funding Agencies: Deutsche Forschungsgemeinschaft (DFG) (Gepris) and National Science Centre, Poland

Project Partner: Center for Research on Prejudice, University of Warsaw http://cbu.psychologia.pl/