# Leveraging Collection-Wide Similarities for Unsupervised Document Structure Extraction Source: [Leveraging Collection-Wide Similarities for Unsupervised Document Structure Extraction]() Collection index: [llms.txt]() Data updated: 2026-06-13T05:00:05.485Z Authors: Gili Lior, Yoav Goldberg, Gabriel Stanovsky Published: 2024-02-21T16:22:21.000Z ## Abstract Document collections of various domains, e.g., legal, medical, or financial, often share some underlying collection-wide structure, which captures information that can aid both human users and structure-aware models. We propose to identify the typical structure of document within a collection, which requires to capture recurring topics across the collection, while abstracting over arbitrary header paraphrases, and ground each topic to respective document locations. These requirements pose several challenges: headers that mark recurring topics frequently differ in phrasing, certain section headers are unique to individual documents and do not reflect the typical structure, and the order of topics can vary between documents. Subsequently, we develop an unsupervised graph-based method which leverages both inter- and intra-document similarities, to extract the underlying collection-wide structure. Our evaluations on three diverse domains in both English and Hebrew indicate that our method extracts meaningful collection-wide structure, and we hope that future work will leverage our method for multi-document applications and structure-aware models. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Document Grounding on Form-10K (grounding gold labels)](): Document Grounding; Form-10K (grounding gold labels); Macro F1 (partial): 93.7. - [Document Grounding on Hebrew Verdict Decisions human-annotated gold labels](): Document Grounding; Hebrew Verdict Decisions human-annotated gold labels; Macro F1 (Partial): 78.1. - [Header Intrusion Detection on Form-10K](): Header intrusion detection; Form-10K; Accuracy: 67.5. - [Header Intrusion Detection on CUAD](): Header intrusion detection; CUAD; Accuracy: 61.7. - [Header Intrusion Detection on Hebrew Verdict Decisions](): Header intrusion detection; Hebrew Verdict Decisions; Accuracy: 65.2. ## Original sources and implementation [Original paper]() [PDF]()