Countering Hate Speech through Prosocial Student Projects

Abstract

We discuss misinformation, conspiracy mongering, and hate speech against vulnerable populations as a threat to social justice, democracy, and world peace. Since the hate in vile statements lies in context rather than on specific words, automated detection is hard. Efforts to automate hate speech detection by traditional methods such as logistic regression, decision trees, random forests and linear SVM as well as by Machine Learning methods are making progress but still fall short. We suggest to counter hate speech by empowering empathy. We advocate engaging students in building prosocial apps for smartphones or tablets to help convey empathy and compassion for those vilified by hate speech. Researchers at ETH and the University of Zurich have found that posting supportive messages for victims of hate speech can counter it more effectively than any other strategy. From our experience teaching ethics and teamwork to upper-level Computer Science and Electrical Engineering students and directing CS capstone projects, we submit that students can be encouraged to develop prosocial, thoughtful apps. Ethical, prosocial software design competitions such as the "Ethics in Action: Art or App Design Challenge"sponsored by the University of Alabama at Birmingham could help.