grobid

A machine learning software for extracting information from scholarly documents

What is grobid?

A machine learning software for extracting information from scholarly documents

Overview of grobid

What it is

GROBID (Generation Of Bibliographic Data) is a machine learning library for extracting, parsing, and re-structuring raw documents such as PDF into structured XML/TEI encoded documents. It focuses on technical and scientific publications. Development began in 2008 and the tool was made open source in 2011. GROBID is designed for production use and is deployed by organizations including ResearchGate, Semantic Scholar, HAL Research Archive, scite.ai, Academia.edu, Internet Archive Scholar, INIST-CNRS, and CERN (Invenio).

Key features

  • Header extraction and parsing from PDF articles, covering title, abstract, authors, affiliations, keywords, etc.
  • References extraction and parsing from PDF articles, achieving around .87 F1-score on an independent PubMed Central set of 1943 PDFs and around .90 on a similar bioRxiv set of 2000 PDFs (using the Deep Learning citation model).
  • Citation contexts recognition and resolution, with accuracy between .76 and .91 F1-score depending on the evaluation collection.
  • Full text extraction and structuring from PDF articles, including document segmentation and structuring of text body (paragraph, section titles, reference and footnote callouts, figures, tables, data availability statements, etc.).
  • PDF coordinates for extracted information, enabling augmented interactive PDF based on bounding boxes.
  • Parsing of references in isolation (above .90 F1-score at instance-level, .95 F1-score at field level using the Deep Learning model).
  • Parsing of names, affiliation and address blocks, dates, and consolidation/resolution of extracted bibliographical references using biblio-glutton or CrossRef REST API (DOI/PMID resolution higher than 0.95 F1-score).
  • Extraction and parsing of patent and non-patent references in patent publications.
  • Extraction of funders and funding information with optional matching against the CrossRef Funder Registry.
  • Identification of copyrights' owner and license associated to the document.

Use cases

GROBID is used for processing large volumes of scholarly PDFs to extract structured metadata and full text. It is suitable for building digital libraries, research information systems, citation databases, and any application requiring automated extraction of bibliographic data, references, and full-text content from scientific publications.

Getting started

See the repository README for detailed installation instructions, including JDK setup and platform-specific requirements. GROBID requires OpenJDK 21 for building from source and runs on Linux (64 bits) or macOS (Intel and ARM). Optional dependencies include Python 3.10-3.11 with JEP for Deep Learning models and NVIDIA GPU with CUDA support.

License and pricing

The license is Apache-2.0. The pricing type is open source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/grobidorggrobid.svg)](https://yunodb.com/tools/grobidorggrobid?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/grobidorggrobid?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/grobidorggrobid.svg" alt="Featured on YunoDB" height="20"></a>