🤖 AI Summary
Detecting discriminatory language in Chinese is highly challenging due to its implicit intent and strong context dependence. This work proposes MAAM, a lightweight, model-agnostic framework that introduces the novel Myopia–Astigmatism anchoring mechanism to preserve semantics relevant to discrimination judgments. MAAM further calibrates predictions by integrating three contextual priors: contextual tone, group identity, and stance polarity (C-I-S). The study also constructs ChLGBT, the first Chinese corpus focused on LGBT-related discriminatory language. Evaluated across various encoder architectures, MAAM consistently achieves substantial improvements in accuracy, F1 score, Brier score, and calibration performance under both zero-shot and few-shot settings, matching the effectiveness of large language models while offering superior compactness and stability.
📝 Abstract
Chinese discriminatory-language detection is challenging because harmful intent is often implicit and context-dependent. We propose MAAM (Myopia--Astigmatism Anchor Mechanism), a lightweight, model-agnostic framework inspired by functional visual blur: rather than preserving every token equally, MAAM retains discrimination-relevant semantic anchors and calibrates them with C--I--S contextual priors (Contextual Tone, Group Identity, and Stance Polarity). We also introduce ChLGBT, to our knowledge the first Chinese LGBT-focused discriminatory-language dataset, with 8,120 manually annotated samples and three ordinal labels: explicit bias, implicit bias, and emotional intensity. Across strong encoder baselines, MAAM improves all three prediction dimensions, with consistent gains in accuracy, F1, Brier score, and expected calibration error. Compared with frontier LLM baselines under zero-shot and few-shot prompting protocols, MAAM remains competitive while offering stronger compactness and stability. These results suggest that interpretable anchor preservation and contextual calibration provide a practical alternative to heavier model scaling for Chinese discriminatory-language assessment.