源保留对齐用于科学PDF中稳健的证据定位

中文摘要 AI 辅助

科学信息抽取系统通常返回带有证据字符串的声明，用户必须在原始PDF中定位该证据。这具有挑战性，因为抽取的证据与PDF文本层是不同表示：换行、Unicode变体、上标、引用标记和碎片化条目会改变文本序列和几何形状。我们提出一个源保留对齐框架：对文本进行归一化以进行稳健匹配，同时保留来源信息以实现准确定位。该框架将证据与归一化后的页面文本对齐，将匹配映射回源字符跨度，并仅渲染其几何形状。当精确对齐失败时，支持换行感知的令牌对齐可恢复支持的跨度，同时排除不匹配的噪声。在1,020篇化学论文上的实验表明，该框架实现了92.6%的引文级自动定位率，而文本搜索为43.6%，预计算边界框基线为19.1%。组件消融实验确认了归一化和近似令牌对齐的各自贡献，而人工验证评估了返回高亮的视觉正确性。总体而言，这些结果表明，可靠的证据验证需要在共享的源保留对齐表示中进行稳健匹配和精确定位。

英文摘要

Scientific information-extraction systems often return a claim with an evidence string, which users must locate in the original PDF. This is challenging because the extracted evidence and PDF text layer are different representations: line wrapping, Unicode variants, superscripts, citation markers, and fragmented items alter text sequences and geometry. We present a source-preserving alignment framework: normalize text for robust matching while preserving provenance for accurate localization. It aligns evidence with normalized page text, maps matches back to source-character spans, and renders only their geometry. When exact alignment fails, line-break-aware token alignment recovers supported spans while excluding unmatched noise. Experiments on 1,020 chemistry papers show that the framework achieves a 92.6\% quote-level automatic localization rate, compared with 43.6\% for text search and 19.1\% for a precomputed bounding-box baseline. Component ablation confirms distinct contributions from normalization and approximate token alignment, while human verification assesses the visual correctness of returned highlights. Overall, these results demonstrate that reliable evidence verification requires robust matching and precise localization within a shared source-preserving alignment representation.

发表机构

• University of Science and Technology of China（中国科学技术大学）
• State Key Laboratory of Cognitive Intelligence（认知智能国家重点实验室）
• CCCC Second Highway Consultants Co., Ltd.（中交第二公路勘察设计研究院有限公司）

机构由 AI 辅助整理，请以论文原文为准。

补充信息

备注

5 pages, 4figures