arXiv cs.CL· Han Wang, Hanqing Tao, Xuesong Wang, Yi Zhu·· 12 小时前AI 评分31
Sledgehammer or Scalpel？FAID：面向隐式仇恨言论检测的细粒度自适应框架
Sledgehammer or Scalpel? A Fine-grained Adaptive Framework for Implicit Hate Speech Detection
AI 导读
研究者提出 FAID 框架，先将隐式仇恨言论细分为 Shallow、Targeted、Context-Dependent 三类再分别适配处理：Shallow 样本用轻量 Prompt-Tuning 快速分类，Targeted 样本用知识增强迭代细化并揭示隐藏目标，Context-Dependent 样本用智能体框架自动生成提示词补全背景信息。
正文 · 原文
Abstract:Unlike explicit attacks with obvious profanity, implicit hate speech hides malice within seemingly compliant expressions through metaphors and contextual hints, making its detection in online content review challenging. Many PLM- or LLM-based methods apply the same processing pipeline across samples, which may overlook differences in information requirements and introduce unnecessary computation. We observe that online hate speech is not monolithic but manifests in varied forms. We therefore define three fine-grained categories: Shallow, Targeted, and Context-Dependent. Accordingly, we propose Fine-grained Adaptive Implicit Hate speech Detection (FAID), a novel framework that first performs fine-grained classification and then adapts to specific categories. Specifically, for Shallow samples with surface-identifiable intents, the framework adopts lightweight Prompt-Tuning for rapid classification; for Targeted comments that bind malicious intent to concealed targets, we design knowledge augmentation to iteratively refine the model and reveal hidden targets; for Context-Dependent comments lacking background information, we utilize an agentic framework that automatically generates prompts to evolve context, infer missing background information and identify ambiguous malicious intents. This adaptive architecture focuses computational resources on complex implicit samples while avoiding redundant reasoning for shallow samples. Experiments on four benchmark datasets demonstrate that FAID significantly outperforms SOTA baselines.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2608.27462 [cs.CL] |
| (or arXiv:2608.27462v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2608.27462 arXiv-issued DOI via DataCite |
Submission history
From: Yi Zhu [view email]
[v1]
Fri, 10 Jul 2026 02:41:55 UTC (637 KB)
[v2]
Mon, 2 Nov 2026 07:45:19 UTC (746 KB)
来源：arXiv cs.CL · arxiv.org