arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.13818cs.CL
收藏

超越仇恨言论:NLP在揭示去人性化语言中的挑战与机遇

Beyond Hate Speech: NLP's Challenges and Opportunities in Uncovering Dehumanizing Language

Politecnico di Milano(米兰理工大学);University of Milan(米兰大学);University College London(伦敦大学学院);University of Sheffield(谢菲尔德大学)

Hamidreza Saffari, Mohammadamin Shafiei, Hezhao Zhang, Lasana Harris, Nafise Sadat Moosavi

AI总结:

本研究系统评估了Claude、GPT、Mistral和Qwen四个大型语言模型在去人性化语言检测任务上的表现,发现仅Claude在优化配置下F1超80%,且模型在不同目标群体上存在系统性预测差异,凸显了群体层面评估的必要性。

摘要:

去人性化,即否认个体或群体的人类特质,是一种特别有害的仇恨言论形式,可使针对边缘化群体的暴力行为正常化。尽管NLP在检测一般仇恨言论方面取得了进展,但由于标注数据稀缺以及此类表达的微妙性质,识别去人性化语言的方法仍然有限。在本研究中,我们系统评估了四个最先进的大型语言模型(LLMs)——Claude、GPT、Mistral和Qwen——在去人性化检测方面的表现。结果表明,在优化配置下,仅有一个模型——Claude——取得了强劲表现(超过80%的F1),而其他模型尽管具备能力,表现仅为中等。当将去人性化与诽谤等相关仇恨类型区分开来时,性能进一步下降。我们还发现目标群体之间存在系统性差异:模型倾向于对某些身份(如Gay men)过度预测去人性化,而对其他身份(如Refugees)则识别不足。这些发现表明,在将预训练语言模型应用于去人性化检测任务时,需要进行系统性的群体层面评估。

英文摘要:

Dehumanization, i.e., denying human qualities to individuals or groups, is a particularly harmful form of hate speech that can normalize violence against marginalized communities. Despite advances in NLP for detecting general hate speech, approaches to identifying dehumanizing language remain limited due to scarce annotated data and the subtle nature of such expressions. In this work, we systematically evaluate four state-of-the-art large language models (LLMs) - Claude, GPT, Mistral, and Qwen - for dehumanization detection. Our results show that only one model-Claude-achieves strong performance (over 80% F1) under an optimized configuration, while others, despite their capabilities, perform only moderately. Performance drops further when distinguishing dehumanization from related hate types such as derogation. We also identify systematic disparities across target groups: models tend to over-predict dehumanization for some identities (e.g., Gay men), while under-identifying it for others (e.g., Refugees). These findings motivate the need for systematic, group-level evaluation when applying pretrained language models to dehumanization detection tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑