CounterGeDi: A controllable approach to generate polite, detoxified and emotional counterspeech
This work addresses the need for more effective and controlled counterspeech generation to assist counter speakers in online hate mitigation, representing an incremental improvement over existing generation models.
The paper tackled the problem of generating counterspeech to combat online hate by proposing CounterGeDi, an ensemble of generative discriminators to guide a DialoGPT model toward more polite, detoxified, and emotional counterspeech, resulting in improvements such as a 15% increase in politeness, 6% in detoxification, and at least 10% in emotion scores across datasets.
Recently, many studies have tried to create generation models to assist counter speakers by providing counterspeech suggestions for combating the explosive proliferation of online hate. However, since these suggestions are from a vanilla generation model, they might not include the appropriate properties required to counter a particular hate speech instance. In this paper, we propose CounterGeDi - an ensemble of generative discriminators (GeDi) to guide the generation of a DialoGPT model toward more polite, detoxified, and emotionally laden counterspeech. We generate counterspeech using three datasets and observe significant improvement across different attribute scores. The politeness and detoxification scores increased by around 15% and 6% respectively, while the emotion in the counterspeech increased by at least 10% across all the datasets. We also experiment with triple-attribute control and observe significant improvement over single attribute results when combining complementing attributes, e.g., politeness, joyfulness and detoxification. In all these experiments, the relevancy of the generated text does not deteriorate due to the application of these controls