Work

Towards a Bulgarian Historical Newspaper Corpus – Construction of Reading Order over the Text in Searchable PDFs

Year: 2026

Type: conference-paper

Abstract: The first task in the process of preparing historical newspaper corpora is to determine the reading order of texts.These texts are often extracted from searchable PDFs produced by some OCR software.In this paper we present an algorithm for generating the original reading order of the text blocks selected from the corresponding PDF.We also performed a tuning of the algorithm parameters.The optimization provides an improvement of 10 %.

Source: Language Resources and Evaluation Conference (LREC)

DOI: 10.63317/336sd7ixv3fw

Authors: Nikolay Paev, Stefan Marinov, Ivan Kratchanov, Petya Osenova, Kiril Simov

Institutions: Bulgarian Academy of Sciences, Institute of Information and Communication Technologies, Ivan Vazov National Library

Language: English

Topic: Handwritten Text Recognition Techniques

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

OpenAlex ID: W7166541280

FWCI:

Cites:

Cited by:

Related to: 10