arXiv cs.CL· Songtao Li, Shidi Zhang, Jianyuan Yuan, Fengyu Zhang·· 12 小时前AI 评分20
GAMA：面向 schema-as-code 生物医学命名实体识别的指南增强多智能体框架
A Guideline-Augmented Multi-Agent Framework for Schema-as-Code Biomedical Named Entity Recognition
AI 导读
GAMA 是一个指南增强的多智能体框架，用于 schema-as-code 生物医学命名实体识别（BioNER）。它先从标注训练样本中归纳候选标注规则并验证，构建数据集专属指南记忆，再由规划组件生成带理由的 span-type 假设、编码组件转为 schema 约束的实体对象，验证模块检查边界、类型与结构合规性并做双循环修正。
正文 · 原文
Abstract:Large language models (LLMs) have shown promising potential for biomedical named entity recognition (BioNER) through instruction following and in-context learning. However, existing LLM-based BioNER methods still face two key limitations. First, retrieved demonstrations and external biomedical knowledge provide limited support for dataset-specific annotation semantics, leaving entity boundaries, type scopes, and annotation conventions ambiguous. Second, free-form generation lacks sufficient structural control, often leading to invalid formats, hallucinated mentions, duplicated entities, and boundary errors. To address these limitations, we propose GAMA, a guideline-augmented multi-agent framework for schema-as-code BioNER. GAMA first induces candidate annotation rules from labeled training instances and verifies them against annotated data to construct reliable dataset-specific guideline memory. Guided by these verified rules, a planning component generates ranked span-type hypotheses with rationales, and a coding component converts them into schema-constrained entity objects. A verification module then checks span grounding, type validity, and structural compliance, and performs dual-loop refinement to correct invalid or low-confidence predictions. Experiments on three widely used BioNER datasets using two different LLMs demonstrate the effectiveness of GAMA, achieving the best performance in four out of six LLM-dataset combinations. Ablation studies further validate the contributions of its key components.
| Comments: | Revised the manuscript based on reviewers' comments and suggestions |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.02970 [cs.CL] |
| (or arXiv:2610.02970v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02970 arXiv-issued DOI via DataCite |
Submission history
From: Songtao Li [view email]
[v1]
Fri, 2 Oct 2026 08:06:20 UTC (1,552 KB)
[v2]
Tue, 6 Oct 2026 08:09:03 UTC (1,552 KB)
[v3]
Fri, 23 Oct 2026 11:18:02 UTC (1,552 KB)
来源：arXiv cs.CL · arxiv.org