超越仇恨：区分多模态内容审核中的不文明与不宽容言论
Beyond Hate: Differentiating Uncivil and Intolerant Speech in Multimodal Content Moderation
AI总结：
本文提出细粒度标注方案区分不文明与不宽容言论，通过2030个仇恨漫画数据验证，细粒度标注提升模型性能，减少有害内容误判。
摘要：
当前多模态毒性基准通常使用单一二元仇恨标签。这种粗略方法混淆了表达的两个根本特征：语气和内容。基于沟通科学理论，我们引入细粒度标注方案，区分两个可分离维度：不文明（粗鲁或 dismissive 语气）和不宽容（攻击多元主义并针对群体或身份的内容），并将其应用于2030个来自仇恨漫画数据集的漫画。我们评估了不同视觉-语言模型在粗标签训练下的表现，跨标签方案的迁移学习以及结合粗仇恨标签与细粒度标注的联合学习方法。我们的结果表明，细粒度标注补充了现有粗标签，并且当联合使用时提高了整体模型性能。此外，使用细粒度方案训练的模型表现出更平衡的审核相关错误分布，并且比仅使用仇恨标签训练的模型更不容易低估有害内容（FNR-FPR，假阴性率与假阳性率之差：LLaVA-1.6-Mistral-7B为0.74至0.42；Qwen2.5-VL-7B为0.54至0.28）。这项工作通过提升数据质量，为内容审核的数据导向方法做出了贡献。总体而言，结合粗细粒度标签为更可靠的多模态审核提供了实用途径。
英文摘要：
Current multimodal toxicity benchmarks typically use a single binary hatefulness label. This coarse approach conflates two fundamentally different characteristics of expression: tone and content. Drawing on communication science theory, we introduce a fine-grained annotation scheme that distinguishes two separable dimensions: incivility (rude or dismissive tone) and intolerance (content that attacks pluralism and targets groups or identities) and apply it to 2,030 memes from the Hateful Memes dataset. We evaluate different vision-language models under coarse-label training, transfer learning across label schemes and a joint learning approach that combines the coarse hatefulness label with our fine-grained annotations. Our results show that fine-grained annotations complement existing coarse labels and, when used jointly, improve overall model performance. Moreover, models trained with the fine-grained scheme exhibit more balanced moderation-relevant error profiles and are less prone to under-detection of harmful content than models trained on hatefulness labels alone (FNR-FPR, the difference between false negative and false positive rates: 0.74 to 0.42 for LLaVA-1.6-Mistral-7B; 0.54 to 0.28 for Qwen2.5-VL-7B). This work contributes to data-centric approaches in content moderation by improving the reliability and accuracy of moderation systems through enhanced data quality. Overall, combining both coarse and fine-grained labels provides a practical route to more reliable multimodal moderation.
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。
