用轻量微调让大模型准确提取文档级事件要素，兼顾角色约束与结构一致性。
Schema-Constrained Document-Level Event Argument Extraction with Lightweight LLM Fine-Tuning
- 通过提示注入角色集确保输出符合事件模式要求
- 在MAVEN-ARG上达到42.39%事件共指F1，超越以往GPT基线
- 仅需少量参数微调，适合实际部署的高效事件抽取
事件要素抽取（EAE）将文档转化为结构化事件记录，识别要素片段并分配预定义角色。文档级EAE因触发词与要素间长程依赖、跨句上下文及严格角色约束，常导致边界错误、角色不确定和模式不一致。本文研究中等规模开源大模型在MAVEN-ARG数据集上是否能可靠完成文档级schema约束型EAE。方法结合三方面：(i) 提示中注入角色集以保证模式合规；(ii) 使用相同JSON接口进行参数高效微调（LoRA）；(iii) 确定性解码后处理，验证JSON格式、过滤无效角色、去重要素并对齐到文档窗口。在官方MAVEN-ARG评估器下，微调后的中等规模模型在提及、实体共指和事件共指任务上均优于此前报告的GPT基线，最佳模型（Phi-4, 14B）在事件共指层面达到42.39% F1。实验代码已公开于https://github.com/dessertlab/EAE/。
原文摘要 · Abstract (English)
Event Argument Extraction (EAE) converts documents into structured event records by identifying argument spans and assigning them schema-defined roles. Document-level EAE is challenging due to long-range dependencies between triggers and arguments, cross-sentence context, and strict role constraints, which often lead to boundary errors, uncertainty in roles, and inconsistencies with restricted schemas. In this paper, we study whether mid-sized open LLMs can perform schema-constrained EAE reliably at the document level on MAVEN-ARG. Our approach combines (i) role-set injection in prompts for schema compliance, (ii) parameter-efficient supervised fine-tuning (LoRA) using the same JSON-only interface used at inference, and (iii) deterministic decoding with post-processing that validates JSON, filters invalid roles, de-duplicates arguments, and aligns spans to the document window. Under the official MAVEN-ARG evaluator, fine-tuned mid-sized open models outperform previously reported GPT baselines across mention, entity-coreference, and event-coreference evaluations; our best model (Phi-4, 14B) reaches 42.39\% F1 at the event-coreference level. Code to reproduce experiments is publicly available at https://github.com/dessertlab/EAE/.
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。