跨语言评估低资源语种偏见，验证高资源方法可迁移
Bias Beyond English: Evaluating Social Bias and Debiasing Methods in a Low-Resource Setting
- 用高资源语料构建多语言偏见评测集，实现跨语言公平比较
- 在5种语言中验证模型在性别、宗教等4维度的偏见表现
- 证明英语偏见缓解方法可有效迁移至中文、泰语等低资源语言
语言模型中的社会偏见可能加剧社会不平等。尽管该问题受到广泛关注，但多数研究集中于英语数据。在低资源场景下，由于训练数据不足，模型性能通常更差。本研究旨在利用高资源语言语料，评估低资源语言中的偏见并实验去偏方法。我们在英语、中文、俄语、印尼语和泰语五种语言上评估了近期多语言模型，并分析了性别、宗教、国籍和种族-肤色四个偏见维度。通过构建多语言偏见评测数据集，本研究实现了不同语言间模型的公平比较。进一步探究了三种去偏方法（CDA、Dropout、SenDeb），结果表明高资源语言中的去偏方法可有效迁移至低资源语言，为多语言NLP公平性研究提供了可操作的洞见。
原文摘要 · Abstract (English)
Social bias in language models can potentially exacerbate social inequalities. Despite it having garnered wide attention, most research focuses on English data. In a low-resource scenario, the models often perform worse due to insufficient training data. This study aims to leverage high-resource language corpora to evaluate bias and experiment with debiasing methods in low-resource languages. We evaluated the performance of recent multilingual models in five languages: English, Chinese, Russian, Indonesian and Thai, and analyzed four bias dimensions: gender, religion, nationality, and race-color. By constructing multilingual bias evaluation datasets, this study allows fair comparisons between models across languages. We have further investigated three debiasing methods-CDA, Dropout, SenDeb-and demonstrated that debiasing methods from high-resource languages can be effectively transferred to low-resource ones, providing actionable insights for fairness research in multilingual NLP.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。