发表机构
University of Pisa; IIT-CNR; ILC-CNR（比萨大学; 意大利国家研究委员会信息科学与技术研究所; 意大利国家研究委员会语言与计算研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对现有通用反言语忽略语境与用户特征的问题，提出情境化反言语生成策略，结合多指标评估发现轻量个性化策略可提升反言语的感知充分性与说服力，为反言语系统开发提供方向。
AI 中文摘要
AI生成的反言语是缓解网络毒性、促进更具建设性对话的可扩展有效策略。但现有方法采用通用的一刀切范式，忽略了对话语境和目标用户特征。本文提出并评估多种生成情境化反言语的策略，使其适配审核场景并个性化目标用户。具体而言，我们探索整合不同形式语境信息与微调技术的多种配置，结合定量指标与预先注册的混合设计众包实验开展全面评估。为确保稳健性，我们基于ROUGE、BLEU和BERTScore实现反言语质量的算法衡量，观察到各指标结果整体一致。此外，我们分析生成的反言语与被审核的有毒消息的哪些特征对感知说服力影响最大，为情境化干预如何更有效提供见解。研究发现个性化策略虽有效但并非统一有效，结合对话语境和用户历史的轻量策略可提升感知充分性与说服力，而其他若干情境化策略则会降低人类感知的反言语质量。综上，这些结果为开发更个性化、有效且负责任的反言语系统提供了可操作方向，最终推进在线内容审核领域的人机协作。
英文摘要
AI-generated counterspeech offers a scalable and effective strategy to mitigate online toxicity by promoting more constructive dialogue. Yet, existing approaches adopt a generic, one-size-fits-all paradigm, overlooking the conversational context and characteristics of the targeted users. Here, we propose and evaluate multiple strategies for generating contextualized counterspeech that is adapted to the moderation setting and personalized to the moderated user. In detail, we explore a range of configurations that integrate different forms of contextual information and fine-tuning techniques. We conduct a comprehensive evaluation combining quantitative indicators with a pre-registered, mixed-design crowdsourcing experiment. To ensure robustness, we implement algorithmic measures of counterspeech quality based on ROUGE, BLEU, and BERTScore, observing overall consistent results across metrics. Furthermore, we analyze which characteristics of both the generated counterspeech and the moderated toxic message most strongly influence perceived persuasiveness, yielding insights into how contextualized interventions can be made more effective. Our findings show that personalization can be effective, but not uniformly so. Lightweight strategies combining conversational context and user history improve perceived adequacy and persuasiveness, whereas several other contextualization strategies degrade human-perceived counterspeech quality. Taken together, these results provide actionable directions for developing more personalized, effective, and responsible counterspeech systems, ultimately advancing human-AI collaboration in online content moderation.
CommentsThis work is an extension of this conference paper: Cima, L., Miaschi, A., Trujillo, A., Avvenuti, M., Dell'Orletta, F., & Cresci, S. (2025, April). Contextualized counterspeech: Strategies for adaptation, personalization, and evaluation. In Proceedings of the ACM on Web Conference 2025 (pp. 5022-5033)
