Handwritten Text Recognition of Ukrainian Manuscripts in the 21st Century: Possibilities, Challenges, and the Future of the First Generic AI-based Model
DOI:
https://doi.org/10.18523/2313-4895.11.2024.226-247Keywords:
Ukrainian, handwritten text recognition, manuscripts, handwriting, AIAbstract
This article reports on developing and evaluating a generic Handwritten Text Recognition (HTR) model created for the automatic computer-assisted transcription of Ukrainian handwriting publicly available via the HTR platform Transkribus. The model’s training process encompasses diverse datasets, including historical manuscripts by renowned poets Taras Shevchenko and Lesya Ukrainka, along with private correspondence used for the General Regionally Annotated Corpus of Ukrainian (GRAC) and a diary procured at the Holodomor Museum collection. We evaluate the model’s performance by comparing its theoretical accuracy, with a character error rate (CER) of 4.2%, against its practical efficacy when augmented with an AI-based language model for Ukrainian and a Large Language Model. The model is versatile and functional and can thus be applied for mass-digitization of Ukrainian cultural heritage. In our outlook section, we identify possibilities for further improving the model.
References
Adobe Acrobat. Easily edit your scanned PDF documents with OCR. Accessed June 14, 2023. https://www.adobe.com/acrobat/how-to/ocr-software-convert-pdf-to-text.html.
Bodnia, Yevhen, & Mariia Kozulia. “Web Application System of Handwritten Text Recognition.” COLINS-2021: 5th International Conference on Computational Linguistics and Intelligent Systems, April 22–23, 2021, Kharkiv, Ukraine. https://ceur-ws.org/Vol-2870/paper98.pdf.
Burlacu, Constanța, & Achim Rabus. “Digitising (Romanian) Cyrillic Using Transkribus: New Perspectives.” Diacronia 14 (December 12, 2021): A196(1–9). https://doi.org/10.17684/i14A196en.
Church Slavonic (2): Free Public AI Model for Handwritten Text Recognition with Transkribus, accessed July 27, 2023, https://readcoop.eu/model/church-slavonic-2/.
GRAC, accessed August 3, 2023. https://uacorpus.org/Kyiv/ua.
Klekovska, Mimoza, Igor Nedelkovski, Vera Stojcevska-Antic, & Dragan Mihajlov. “Automatic Letter Style Recognition of Churchslavic Manuscripts.” In Proceedings of Papers of the 44th International Scientific Conference on Information, Communication and Energy Systems and Technologies (ICEST 2009), Veliko Tarnovo, Bulgaria, June 25–27, 2009. Vol. 1, 221–4 (Sofia, 2009).
Kornienko, Sergei, Fedor Cherepanov, & Leonid Iasnitckii. “Raspoznavanie tekstov rukopisnykh i staropechatnykh knig na osnove neirosetevykh tekhnologii” [“OCR of manuscripts and early printed books using neural networks”]. Paper presented at conference “Modern Information Technologies and Written Heritage: from Ancient Texts to Electronic Libraries” – El’Manuscript-08, Kazan, Republic of Tatarstan, August 25–30, 2008. https://textualheritage.org/ru/el-manusctipt-08-/52.html.
Klyment, Kvitka, ed. Narodni melodii. Z holosu Lesi Ukrainky [Folk Melodies. From the Voice of Lesya Ukrainka]. Vol. 1. Kyiv, 1917.
Martinovska, Cveta, Mimoza Klekovska, Igor Nedelkovski, & Dragan Kaevski. “Methodologies for Recognition of Old Slavic Cyrillic Characters.” International Journal of Computational Intelligence Studies 2, no. 3–4 (January 2013): 264–87. https://doi.org/10.1504/IJCISTUDIES.2013.057639.
Muehlberger, Guenter et al. “Transforming Scholarship in the Archives through Handwritten Text Recognition: Transkribus as a Case Study.” Journal of Documentation 75, no. 5 (2019): 954–76. https://doi.org/10.1108/JD-07-2018-0114.
MultiHTR – Multilinguale Handschriftenerkennung. Projektbeschreibung. Аccessed July 27, 2023. https://www.multihtr.uni-freiburg.de.
Namboodiri, Anoop M., & Anil K. Jain. “Online Handwritten Script Recognition.” IEEE Transactions on Pattern Analysis and Machine Intelligence 26, no. 1 (January 2004): 124–30. https://doi.org/10.1109/TPAMI.2004.1261096.
Optical Character Recognition (and more) for everyone, accessed June 14, 2023, https://www.ocr4all.org/.
Rabus, Achim. “Recognizing Handwritten Text in Slavic Manuscripts: A Neural-Network Approach Using Transkribus’. Scripta & E-Scripta 19 (2019): 9–32. 246 Kyiv-Mohyla Humanities Journal 11 (2024)
Ridni: henealohichne tovarystvo. Doslidzhennia rodovodu v Ukraini [Ridni: genealogical society. Research on genealogy in Ukraine]. Аccessed: July 27, 2023, https://ridni.org/.
Savic, M. D., & M. Bojovic. “Recognition of Handwritten Text: Basic Concepts of a New Approach.” In 4th International Conference on Telecommunications in Modern Satellite, Cable and Broadcasting Services. TELSIKS’99 (Cat. No.99EX365), 2:468–71, vol. 2, 1999. https://doi.org/10.1109/TELSKS.1999.806253.
Saving Ukrainian Cultural Heritage Online. Accessed August 3, 2023, https://www.sucho.org/.
Scripta / escriptorium. GitLab, accessed July 27, 2023. https://gitlab.com/scripta/escriptorium.
Shevchenko, Taras. “Posadzhu kolo khatyny...”: virsh, rukopys, bilovyi avtohraf ostannoi redaktsii virsha “Podrazhaniie” v rukopysnomu zbirnyku P. O. Kulisha [“I Will Plant Near the Hut...”: poem, manuscript, clean autograph of the last edition of the poem “Imitation” in the manuscript collection of P. O. Kulish]. 187?-188? No. 28438. Fond I: Literaturni materialy [Literary materials]. Manuscript Institute of the V. I. Vernadskyi National Library, Kyiv. http://irbis-nbuv.gov.ua/dlib/item/0000613.
Shevchenko, Taras. “Topolia” ta inshi virshi. Zbirnyk poezii. Rukopysnyi spysok [“Topolia” and other poems. A collection of poems. Handwritten list]. 18?? No. 7448. Fond I: Literaturni materialy [Literary materials]. Manuscript Institute of the V. I. Vernadskyi National Library, Kyiv.
Tikhonov, Aleksej, Lesley Loew, Milanka Matić-Chalkitis, Martin Meindl, & Achim Rabus. “Multilingual Handwritten Text Recognition (MultiHTR) or Reading Your Grandma’s Old Letters in German, Russian, Serbian, and Ottoman Turkish with Artificial Intelligence.” In The Palgrave Handbook of Digital and Public Humanities, edited by Anne Schwan and Tara Thomson, 215–33. Cham: Springer International Publishing, 2022. https://doi.org/10.1007/978-3-031-11886-9_12.
Ukrainian generic handwriting: Free Public AI Model for Handwritten Text Recognition with Transkribus. Accessed July 27, 2023. https://readcoop.eu/model/ukrainian-generichandwriting/.
Ukrainian language for ABBYY FineReader Professional Edition 8.0.1126.0. Accessed June 14, 2023. https://ukrainian-language-for-abbyy-finereader-professional-edition.updatestar.com/.
Unlock the past with Transkribus. Accessed July 27, 2023. https://www.transkribus.org.
What is the difference between OCR (Optical Character Recognition) and HTR? https://readcoop.eu/transkribus/help/what-is-the-difference-between-ocr-optical-characterrecognition-and-htr/.
Winkler, Alexander. “OCR4All Tools (Cyrillic).” HTML. 2020. Reprint, GitHub, 18 May 2021. https://github.com/alexander-winkler/ocr4all_tools.
Zelenskyi, Volodymyr. “Vziav uchast u konferentsii v Haazi, meta yakoi – poriatunok mizhnarodnoho prava...” [“Participated in a conference in The Hague, the goal of which was to save international law...”]. July 14, 2022. https://t.me/V_Zelenskiy_official/2534. Аccessed: July 25, 2023.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2024 Aleksej Tikhonov, Achim Rabus
This work is licensed under a Creative Commons Attribution 4.0 International License.
Kyiv-Mohyla Humanities Journal provides free access to original research without restriction barriers (i.e. subscription fees, licensing fees etc.). The journal allows re-use of content for non-commercial/educational purposes indexing the source.
Unless otherwise indicated, content is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license, which means you are free to:
copy
distribute
transmit
adapt
and make commercial use of the work
...provided that any use is made with attribution to author(s) and Kyiv-Mohyla Humanities Journal.
The author passes copyright of the article to the journal and Kyiv-Mohyla Humanities Journal; author can archive post-print articles (PDF versions) on s/he web-site (http://www.sherpa.ac.uk).
