Document Information Extraction

Overview

Document information extraction studies how to convert documents into structured fields, entities, relations, events, and tables that can support search, analysis, and decision-making. Common approaches combine OCR and layout analysis with supervised sequence taggers, schema-guided language models, multimodal models, retrieval, and deterministic validation. Work spans receipts and forms as well as long clinical, financial, legal, and scientific documents, with benchmarks increasingly testing noisy inputs, layout understanding, source grounding, and output completeness. A central challenge is balancing flexible semantic interpretation with exact values, boundaries, and reliable evidence, while keeping systems efficient and robust across document types.

Latest papers

CardsList

305 papers21 papers in the last four weeks, down 38% on the four weeks before. 0.2% of all new papers.

  1. Building verifiable databases at scale from the chemical literature via a multi-agent system

    Jul 27, 2025Yufan Chen, Ching Ting Leung, Bowen Yu +6Document Information ExtractionMulti-Agent LLM Systems

  2. Beyond Isolated Dots: Benchmarking Structured Table Construction as Deep Knowledge Extraction

    Jul 22, 2025Tianyun Zhong, Guozhao Mo, Yanjiang Liu +9Document Information ExtractionLLM Evaluation

  3. POSID: Target-Task-Unsupervised Person-Attribute Extraction from Heterogeneous Public-Safety Text

    Jul 5, 2025KMA SolaimanDocument Information Extraction

  4. Business as Rulesual: A Benchmark and Framework for Business Rule Flow Modeling with LLMs

    May 24, 2025Chen Yang, Ruping Xu, Ruizhe Li +2LLM GroundingLLM Prompting

  5. Auto-AdvER: Disentangling Condition, History, and Commercial Offers in Automotive Advertisements

    Dec 7, 2024Filippos Ventirozos, Ioanna Nteka, Tania Nandy +4Document Information ExtractionProduct Attribute Extraction