通过LLM辅助MIPVU规则脚本生成实现可解释的中文隐喻识别：一种比较性协议研究
Interpretable Chinese Metaphor Identification via LLM-Assisted MIPVU Rule Script Generation: A Comparative Protocol Study
AI总结：
本文通过LLM辅助生成可解释的规则脚本，对中文隐喻识别进行跨协议比较，发现协议选择是影响识别结果的主要因素。
摘要：
隐喻识别是隐喻语言处理中的基础任务，然而大多数计算方法都作为不透明的分类器运作，无法解释为何一个表达被判定为隐喻。这种可解释性缺口在中文中尤为突出，因为中文具有丰富的隐喻传统，缺乏形态学线索，且标注资源有限，加大了挑战。我们提出了一种LLM辅助的流程，将四个隐喻识别协议——MIP/MIPVU词汇分析、CMDAG概念映射标注、基于情感的检测以及以明喻为导向的识别——作为可执行的人工可审规则脚本。每个协议都是由确定性步骤组成的模块化链，其中穿插受控的LLM调用，产生结构化的理由，与每一个分类决定一起。我们在七个中文隐喻数据集中进行评估，涵盖词级、句子级和跨度级的标注，建立了中文隐喻识别的第一个跨协议比较。在协议内部评估中，协议A（MIP）在词级识别中达到F1值为0.472，而跨协议分析揭示出显著的分歧：协议A和D之间的配对Cohen's kappa仅为0.001，而协议B和C显示出几乎完美的一致（kappa=0.986）。可解释性审计显示所有协议都实现了100%的确定性可重复性，理由的正确性从0.40到0.87，可编辑性从0.80到1.00。错误分析发现概念-领域不匹配和语域敏感性是主要的失败模式。我们的结果表明，协议选择是隐喻识别中最大的变异性来源，超过模型层面的变异性，并且规则脚本架构在保持完全透明的同时实现了竞争性的性能。
英文摘要：
Metaphor identification is a foundational task in figurative language processing, yet most computational approaches operate as opaque classifiers offering no insight into why an expression is judged metaphorical. This interpretability gap is especially acute for Chinese, where rich figurative traditions, absent morphological cues, and limited annotated resources compound the challenge. We present an LLM-assisted pipeline that operationalises four metaphor identification protocols--MIP/MIPVU lexical analysis, CMDAG conceptual-mapping annotation, emotion-based detection, and simile-oriented identification--as executable, human-auditable rule scripts. Each protocol is a modular chain of deterministic steps interleaved with controlled LLM calls, producing structured rationales alongside every classification decision. We evaluate on seven Chinese metaphor datasets spanning token-, sentence-, and span-level annotation, establishing the first cross-protocol comparison for Chinese metaphor identification. Within-protocol evaluation shows Protocol A (MIP) achieves an F1 of 0.472 on token-level identification, while cross-protocol analysis reveals striking divergence: pairwise Cohen's kappa between Protocols A and D is merely 0.001, whereas Protocols B and C exhibit near-perfect agreement (kappa = 0.986). An interpretability audit shows all protocols achieve 100% deterministic reproducibility, with rationale correctness from 0.40 to 0.87 and editability from 0.80 to 1.00. Error analysis identifies conceptual-domain mismatch and register sensitivity as dominant failure modes. Our results demonstrate that protocol choice is the single largest source of variation in metaphor identification, exceeding model-level variation, and that rule-script architectures achieve competitive performance while maintaining full transparency.
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。
