通过分权审计机制提升跨维基表格的溯源准确性
Stage-Audit: Auditable Source-Frontier Discovery for Cross-Wiki Tables
- 分离撰写与审计权限，每行数据需经源引用审核
- 在15个领域51个实例上精确率提升42%，F1提升35%
- 适合需要可追溯、高可信度数据构建的场景
LLM生成的表格看似有来源支持，实则可能包含无据可查的条目：编辑者可能从参数化记忆中调取信息，并事后添加不真实的页面级引用。本文研究这一风险在Seed2Frontier任务中的表现——从种子页面发现补充页面以构建结构化表格。Stage-Audit采用分离撰写与审计权限、行级源引用闸门，以及涵盖键、模式、来源角色、基数和范围的12项审计分类体系。在覆盖15个顶级域名的51个实例评估集上，相比原始LLM编辑器，Stage-Audit将源前沿精确率从0.356提升至0.505（相对提升42%），F1从0.334升至0.451（相对提升35%），同时保持逐行可追溯的来源记录。该对比凸显了政策设计的贡献，而非单纯基于LLM的发现能力。
原文摘要 · Abstract (English)
LLM-curated tables can appear source-grounded while containing unsupported rows: the curator may recall entries from parametric memory and retroactively attach page-level citations that are not the actual source. We study this hazard in Seed2Frontier discovery: the task of finding complement Wikipedia pages from a seed page to assemble a structured table. Stage-Audit addresses it with disjoint curator-auditor write rights, a row-level source-citation gate, and a 12-check audit taxonomy over keys, schema, source roles, cardinality, and scope. On a curated 51-instance Seed2Frontier evaluation set spanning 15 top-level domains, Stage-Audit improves source-frontier precision over a vanilla LLM curator from 0.356 to 0.505 (+42% relative) and F1 from 0.334 to 0.451 (+35%), while maintaining explicit per-row source traceability. The vanilla-LLM-vs-Stage-Audit comparison isolates the policy contribution rather than LLM-based discovery in general.
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。