Counter Narrative Generation through Knowledge-Injected Prompt Learning

This paper presents a novel method for generating effective counter-narratives to combat hate speech using knowledge-injected prompt learning. Our approach leverages external knowledge sources to create more contextually appropriate and persuasive counter-arguments.

The proposed framework combines prompt learning techniques with external knowledge injection to generate high-quality counter-narratives that can effectively address hate speech while maintaining persuasiveness and contextual appropriateness. Experimental results on Korean and English datasets demonstrate the effectiveness of our approach in generating diverse and effective counter-narratives.

Recommended citation: Hyeonwoo Jung, Byounghan Lee, Kyung-Ah Sohn. (2024). "Counter Narrative Generation through Knowledge-Injected Prompt Learning." Korea Software Congress (KSC2024), Yeosu, South Korea.