Sub-field · 6 papers
LLM Scientific Document Information Extraction
Extracting structured information from scientific documents—including tables, figures, text, and references—using large language models and related AI techniques is the central problem addressed across these papers. Methods span LLM-based information extraction, active learning, uncertainty quantification, and question generation, applied to domains such as toxicology, digital libraries, and archival research. A recurring theme is automating or scaffolding the labor-intensive process of converting unstructured scholarly content into usable, structured data.
Papers in this community
- Structured information extraction from scientific text with large language models 2024 DOI ↗
- Scientific Table Data Extraction with Uncertainty Quantification 2024 DOI ↗
- Scaffolding Inquiry-Oriented Web Search using LLM-based Question Generation 2025 DOI ↗
- LLM-Based Active Learning for Identifying References to Archival Repositories 2025 DOI ↗
- JCDL 2024 Tutorial: Academic Table and Figure Understanding for Digital Libraries 2024 DOI ↗
- Automating Data Extraction From Scientific Literature and General PDF Files Using Large Language Models and KNIME: An Application in Toxicology 2025 DOI ↗