The Dream of a Universal Variorum

Heslin, Peter

doi:10.1093/acprof:oso/9780199688982.003.0025

Cited by 2 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the evaluation set, we selected a sample of five pages (2081±621 words) from each commentary. 9 We deliberately decided against random sampling to ensure that this sample is representative of the different layout types (e.g. introduction, table of contents, commentary, etc.)…”

Section: Ocr Gt For Historical Commentariesmentioning

confidence: 99%

“…Commentaries on Classical Greek and Latin literature present the typical difficulties of historical OCR; they have complex layouts, often with multiple columns and rows of text, and additionally contain Latin and Greek scripts mixed together within the same page and often even within the same line of text. As a result, the OCR quality of commentaries available in open repositories such as the Internet Archive renders these materials inadequate, most of the times, for use in research contexts, despite their recognized potential [9].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optical Character Recognition of 19th Century Classical Commentaries: the Current State of Affairs

Romanello¹,

Najem-Meyer²,

Robertson³

2021

Preprint

View full text Add to dashboard Cite

Together with critical editions and translations, commentaries are one of the main genres of publication in literary and textual scholarship, and have a century-long tradition. Yet, the exploitation of thousands of digitized historical commentaries was hitherto hindered by the poor quality of Optical Character Recognition (OCR), especially on commentaries to Greek texts. In this paper, we evaluate the performances of two pipelines suitable for the OCR of historical classical commentaries. Our results show that Kraken + Ciaconna reaches a substantially lower character error rate (CER) than Tesseract/OCR-D on commentary sections with high density of polytonic Greek text (average CER 7% vs. 13%), while Tesseract/OCR-D is slightly more accurate than Kraken + Ciaconna on text sections written predominantly in Latin script (average CER 8.2% vs. 8.4%). As part of this paper, we also release GT4HistComment, a small dataset with OCR ground truth for 19 th classical commentaries and Pogretra, a large collection of training data and pre-trained models for a wide variety of ancient Greek typefaces.

show abstract

Section: Ocr Gt For Historical Commentariesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%