Features of parsing bibliographies from PDF files of scientific articles
V. I. Tumanov, Tatyana S. Katermina
Source record
Source: Crossref
Published: Aug 31, 2026
DOI: 10.24147/2222-8772.2026.2.100-109
Open original source ↗Source abstract
This article addresses the problem of automatic extraction of bibliographic references from PDF files of scientific publications. It examines major bibliographic citation standards - GOST R 7.0.5-2008, APA, IEEE, MLA, Chicago, and Vancouver - and discusses their specific application to electronic sources. The study analyzes key parsing challenges arising from layout peculiarities, such as two-column formats, breaks in reference entries at page and column boundaries, and multi-page reference lists. Classical NLP approaches (regular expressions, Conditional Random Fields, and specialized systems like GROBID, CERMINE, and AnyStyle) are systematized. The methodology and structure of a comparative experiment involving four systems are presented: Regex + pdfplumber, CERMINE, AnyStyle, GROBID.
Evidence graph
No public relationships recorded yet.
Integrity note: This page is a factual metadata record created by deterministic ingestion. It is not a claim that the work moves a mathematical frontier or has been independently verified.