Counterspeech the ultimate shield! Multi-Conditioned Counterspeech Generation through Attributed Prefix Learning

TL;DR

This paper introduces HiPPrO, a hierarchical prefix learning framework that enhances counterspeech generation by considering multiple attributes simultaneously, leading to more relevant and effective responses against hate speech.

Contribution

The paper proposes a novel two-stage hierarchical prefix learning framework, HiPPrO, that incorporates multiple attributes for improved counterspeech generation, extending previous single-attribute models.

Findings

38% improvement in intent conformity

Enhanced relevance and appropriateness of counterspeech

Open-sourced code and dataset for reproducibility

Abstract

Counterspeech has proven to be a powerful tool to combat hate speech online. Previous studies have focused on generating counterspeech conditioned only on specific intents (single attributed). However, a holistic approach considering multiple attributes simultaneously can yield more nuanced and effective responses. Here, we introduce HiPPrO, Hierarchical Prefix learning with Preference Optimization, a novel two-stage framework that utilizes the effectiveness of attribute-specific prefix embedding spaces hierarchically optimized during the counterspeech generation process in the first phase. Thereafter, we incorporate both reference and reward-free preference optimization to generate more constructive counterspeech. Furthermore, we extend IntentCONANv2 by annotating all 13,973 counterspeech instances with emotion labels by five annotators. HiPPrO leverages hierarchical prefix optimization to integrate these dual attributes effectively. An extensive evaluation demonstrates that HiPPrO achieves a ~38 % improvement in intent conformity and a ~3 %, ~2 %, ~3 % improvement in Rouge-1, Rouge-2, and Rouge-L, respectively, compared to several baseline models. Human evaluations further substantiate the superiority of our approach, highlighting the enhanced relevance and appropriateness of the generated counterspeech. This work underscores the potential of multi-attribute conditioning in advancing the efficacy of counterspeech generation systems. Our code is available on Github and dataset is open-sourced on Hugging-face.