Alignment of Historical Manuscript Transcriptions and Translations

Maarten Janssen, Piroska Lendvai, Anna Jouravel


Abstract
Using an XML-based framework, we compiled a gold standard for alignments in five primary as well as derived texts, related to De Lepra ad Sistelium by Methodius Olympius. These comprise diplomatic transcripts, editions, and translations of this work, involving both historical and modern languages. Using the TEITOK corpus platform, we created sentence-level gold standard alignments for our parallel resp. comparable texts, and applied both neural and classical alignment methods (SentenceBERT, Hunalign, Awesome-Align). We evaluated the methods in terms of Alignment Error Rate. We show that for alignment of our historical texts, Hunalign performs better than deep learning based methods.
Anthology ID:
2025.ranlp-1.55
Volume:
Proceedings of the 15th International Conference on Recent Advances in Natural Language Processing - Natural Language Processing in the Generative AI Era
Month:
September
Year:
2025
Address:
Varna, Bulgaria
Editors:
Galia Angelova, Maria Kunilovskaya, Marie Escribe, Ruslan Mitkov
Venue:
RANLP
SIG:
Publisher:
INCOMA Ltd., Shoumen, Bulgaria
Note:
Pages:
462–470
Language:
URL:
https://aclanthology.org/2025.ranlp-1.55/
DOI:
Bibkey:
Cite (ACL):
Maarten Janssen, Piroska Lendvai, and Anna Jouravel. 2025. Alignment of Historical Manuscript Transcriptions and Translations. In Proceedings of the 15th International Conference on Recent Advances in Natural Language Processing - Natural Language Processing in the Generative AI Era, pages 462–470, Varna, Bulgaria. INCOMA Ltd., Shoumen, Bulgaria.
Cite (Informal):
Alignment of Historical Manuscript Transcriptions and Translations (Janssen et al., RANLP 2025)
Copy Citation:
PDF:
https://aclanthology.org/2025.ranlp-1.55.pdf