Elamur Research⌕
☰
home / papers / 2026 / UppsalaNLP at EvaLatin 2026: Multilingual parsing for Latin

UppsalaNLP at EvaLatin 2026: Multilingual parsing for Latin

Summary

This paper describes Uppsala University's submission to the EvaLatin 2026 competition, which focused on parsing the syntactic dependencies of Latin prose and poetry. Instead of building a new architecture, the author tests how far an existing parser can go when the only change is the choice of training data. That setup is realistic for many digital humanities projects. The MaChamp toolkit, which fine-tunes XLM-RoBERTa Large with a biaffine graph-based parser, is trained on eight UD-compatible Latin treebanks. The experiments compare UD 2.17 versions with harmonized ones, particularly for PROIEL. Further runs add treebanks from other historical languages, optionally leaving out non-Latin PROIEL data, and test multitask training on UPOS, morphology and lemmas. No gold tags are used as input. Additional languages give only modest improvements on the treebanks' own test sets but help more on the shared task data, especially poetry. The primary system wins most prose metrics but ranks only fourth on poetry. A variant further fine-tuned on Latin alone does better there, which highlights genre adaptation as an open problem.

Key points

  • Training data:
    • Latin data includes UD_Circse, ITTB, LLCT, Perseus, PROIEL (harmonized "H-PROIEL") and UDante, plus the UD-compatible Archimedes Latinus and Sabellicus treebanks.
    • Extra languages come in three groups: Ancient Greek; Romance (Old French, Old Occitan); and Other (Gothic, Old Church Slavonic).
  • Parser setup: MaChamp with XLM-RoBERTa Large, a CLE-based non-projective parser, and the dataset sampling smoothing parameter set to 0.5. Results are averaged over three random seeds.
  • Submitted systems:
    • The primary system is trained on Latin plus other historical languages, excluding non-Latin PROIEL treebanks.
    • The secondary system is the same model further trained only on Latin.
  • Official test sets: prose (203 sentences) and poetry (308 sentences). Metrics are LAS and CLAS, each with and without relation subtypes.
  • Rankings:
    • Prose: the primary system is first on most metrics, and the secondary system places second to third.
    • Poetry: the primary system places fourth, and the secondary system places third.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages · Natural language processing · History & humanities

Topics: Multilingual & cross-lingual NLP · Latin

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/57u2jdmxj3ut

@inproceedings{stymne2026uppsalanlp,
  title = {{UppsalaNLP at EvaLatin 2026: Multilingual parsing for Latin}},
  author = {Stymne, Sara},
  booktitle = {Proceedings of the Fourth Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA 2026) @ LREC 2026},
  year = {2026},
  doi = {10.63317/57u2jdmxj3ut},
  url = {https://doi.org/10.63317/57u2jdmxj3ut}
}

References

41 references, 8 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. David Bamman and Patrick J. Burns. 2020. Latin BERT: A contextual language model for classical philology. CoRR, abs/2009.10053. ↗
  2. David Bamman and Gregory Crane. 2011. The Ancient Greek and Latin dependency treebanks. In Language Technology for Cultural Heritage, pages 79–98. Springer Berlin Heidelberg, Berlin, Heidelberg. ↗
  3. Rufus Behr. 2024. Behr at EvaLatin 2024: Latin dependency parsing using historical sentence embeddings. In Proceedings of the Third Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA) @ LREC-COLING-2024, pages 198–202, Torino, Italia. ELRA and ICCL. find
  4. Flavio Massimiliano Cecchini, Timo Korkiakangas, and Marco Passarotti. 2020a. A new Latin treebank for Universal Dependencies: Charters between Ancient Latin and Romance languages. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 933– 942, Marseille, France. European Language Resources Association. find
  5. Flavio Massimiliano Cecchini, Rachele Sprugnoli, Giovanni Moretti, and Marco Passarotti. 2020b. UDante: First steps towards the universal dependencies treebank of Dante’s Latin works. In Proceedings of the Seventh Italian Conference on Computational Linguistics, pages 1–7. Italian Association for Computational Linguistics (AILC). find
  6. Yoeng-Jin Chu and Tseng-Hong Liu. 1965. On the shortest arborescence of a directed graph. Scientia Sinica, 14:1396–1400. find
  7. Alexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2020. Unsupervised cross-lingual representation learning at scale. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 8440–8451, Online. Association for Computational Linguistics. ↗
  8. Vera Danilova and Sara Stymne. 2023. UD-MULTIGENRE – a UD-based dataset enriched with instance-level genre annotations. In Proceedings of the 3rd Workshop on Multi-lingual Representation Learning (MRL), pages 253–267, Singapore. Association for Computational Linguistics. ↗
  9. Miryam de Lhoneux, Sara Stymne, and Joakim Nivre. 2017. Arc-hybrid non-projective dependency parsing with a static-dynamic oracle. In Proceedings of the 15th International Conference on Parsing Technologies, pages 99–104, Pisa, Italy. Association for Computational Linguistics. find
  10. Marie-Catherine de Marneffe, Christopher D. Manning, Joakim Nivre, and Daniel Zeman. 2021. Universal Dependencies. Computational Linguistics, 47(2):255–308. find our card
  11. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 4171–4186, Minneapolis, Minnesota. Association for Computational Linguistics. ↗
  12. Timothy Dozat and Christopher D. Manning. 2018. Simpler but more accurate semantic dependency parsing. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 484–490, Melbourne, Australia. Association for Computational Linguistics. find our card
  13. Federica Gamba and Daniel Zeman. 2023a. Latin morphology through the centuries: Ensuring consistency for better language processing. In Proceedings of the Ancient Language Processing Workshop, pages 59–67, Varna, Bulgaria. IN-COMA Ltd., Shoumen, Bulgaria. find
  14. Federica Gamba and Daniel Zeman. 2023b. Universalising Latin Universal Dependencies: a harmonisation of Latin treebanks in UD. In Proceedings of the Sixth Workshop on Universal Dependencies (UDW, GURT/SyntaxFest 2023), pages 7–16, Washington, D.C. Association for Computational Linguistics. find
  15. Dag Trygve Truslew Haug and Marius Jøhndal. 2008. Creating a parallel treebank of the Old Indo-European Bible translations. In Proceedings of the Second Workshop on Language Technology for Cultural Heritage Data (LaTeCH 2008), pages 27–34. find
  16. Federica Iurescia, Marco Passarotti, and Rachele Sprugnoli. 2026. Overview of the Dependency Parsing Task at EvaLatin 2026. In Proceedings of the Fourth Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA 2026), Palma, Mallorca (Spain). ELRA. find
  17. Antonia Karamolegkou and Sara Stymne. 2021. Investigation of transfer languages for parsing Latin: Italic branch vs. Hellenic branch. In Proceedings of the 23rd Nordic Conference on Computational Linguistics (NoDaLiDa), pages 315– 320, Reykjavik, Iceland (Online). Linköping University Electronic Press, Sweden. find
  18. Hanna-Mari Kristiina Kupari, Erik Henriksson, Veronika Laippala, and Jenna Kanerva. 2024. Improving Latin dependency parsing by combining treebanks and predictions. In Proceedings of the 4th International Conference on Natural Language Processing for Digital Humanities, pages 216–228, Miami, USA. Association for Computational Linguistics. find
  19. Jack Edmonds. 1967. Optimum branchings. Journal of Research of the national Bureau of Standards B, 71(4):233–240. find
  20. Wouter Mercelis. 2024. KU Leuven / Brepols-CTLO at EvaLatin 2024: Span extraction approaches for Latin dependency parsing. In Proceedings of the Third Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA) @ LREC-COLING-2024, pages 203–206, Torino, Italia. ELRA and ICCL. find
  21. Margherita Fantoli and Miryam de Lhoneux. 2022. Linguistic annotation of neo-Latin mathematical texts: A pilot-study to improve the automatic parsing of the archimedes latinus. In Proceedings of the Second Workshop on Language Technologies for Historical and Ancient Languages, pages 129–134, Marseille, France. European Language Resources Association. find
  22. Irene Miani, Sara Stymne, and Gregory R. Darwin. 2026. Cross-lingual and cross-domain transfer learning for POS tagging in historical Germanic low-resource languages. In Proceedings of the Second Workshop on Language Models for Low-Resource Languages (LoResLM 2026), pages 542–558, Rabat, Morocco. Association for Computational Linguistics. find our card
  23. Max Müller-Eberstein, Rob van der Goot, and Barbara Plank. 2021. Genre as weak supervision for cross-lingual dependency parsing. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 4786–4802, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics. ↗
  24. Aaron Smith, Bernd Bohnet, Miryam de Lhoneux, Joakim Nivre, Yan Shao, and Sara Stymne. 2018. 82 treebanks, 34 models: Universal Dependency parsing with multi-treebank models. In Proceedings of the CoNLL 2018 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies, pages 113–123, Brussels, Belgium. Association for Computational Linguistics. find our card
  25. Sebastian Nehrdich and Oliver Hellwig. 2022. Accurate dependency parsing and tagging of Latin. In Proceedings of the Second Workshop on Language Technologies for Historical and Ancient Languages, pages 20–25, Marseille, France. European Language Resources Association. find
  26. Rachele Sprugnoli, Federica Iurescia, and Marco Passarotti. 2024. Overview of the EvaLatin 2024 evaluation campaign. In Proceedings of the Third Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA) @ LREC-COLING-2024, pages 190–197, Torino, Italia. ELRA and ICCL. find
  27. Joakim Nivre, Marie-Catherine de Marneffe, Filip Ginter, Jan Hajič, Christopher D. Manning, Sampo Pyysalo, Sebastian Schuster, Francis Tyers, and Daniel Zeman. 2020. Universal Dependencies v2: An evergrowing multilingual treebank collection. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 4034–4043, Marseille, France. European Language Resources Association. find our card
  28. Joakim Nivre and Chiao-Ting Fang. 2017. Universal Dependency evaluation. In Proceedings of the NoDaLiDa 2017 Workshop on Universal Dependencies (UDW 2017), pages 86–95, Gothenburg, Sweden. Association for Computational Linguistics. find
  29. Marco Passarotti. 2019. The project of the Index Thomisticus treebank. Digital Classical Philology, 10:299–320. find
  30. Nils Reimers and Iryna Gurevych. 2019. Sentence-BERT: Sentence embeddings using Siamese BERT-networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 3982–3992, Hong Kong, China. Association for Computational Linguistics. ↗
  31. Frederick Riemenschneider and Anette Frank. 2023. Exploring large language models for classical philology. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 15181– 15199, Toronto, Canada. Association for Computational Linguistics. find
  32. Natasha Romanova, Rayan Ziane, and Barbara Francioni. 2025. Adaptation of models for parsing of Old Gascon. In Proceedings of Journées scientifiques du réseau thématique LIFT2 linguistique informatique, formelle et de terrain, Paris, France. find
  33. Achim Stein and Sophie Prévost. 2013. Syntactic annotation of medieval texts: The Syntactic Reference Corpus of Medieval French (SRCMF). In Paul Bennett, Martin Durrell, Silke Scheible, and Richard J. Whitt, editors, New Methods in Historical Corpora, Corpus Linguistics and International Perspectives on Language, pages 275–282. Gunter Narr Verlag. find
  34. Milan Straka and Jana Straková. 2017. Tokenizing, POS tagging, lemmatizing and parsing UD 2.0 with UDPipe. In Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies, pages 88–99, Vancouver, Canada. Association for Computational Linguistics. ↗ our card
  35. Milan Straka, Jana Straková, and Federica Gamba. 2024. ÚFAL LatinPipe at EvaLatin 2024: Morphosyntactic analysis of Latin. In Proceedings of the Third Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA) @ LREC-COLING-2024, pages 207–214, Torino, Italia. ELRA and ICCL. find
  36. Sara Stymne. 2020. Cross-lingual domain adaptation for dependency parsing. In Proceedings of the 19th International Workshop on Treebanks and Linguistic Theories, pages 62–69, Düsseldorf, Germany. Association for Computational Linguistics. find
  37. Daniel G. Swanson, Bryce D. Bussert, and Francis Tyers. 2024. Producing a parallel Universal Dependencies treebank of Ancient Hebrew and Ancient Greek via cross-lingual projection. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 13074–13078, Torino, Italia. ELRA and ICCL. find our card
  38. Rob van der Goot and Miryam de Lhoneux. 2021. Parsing with pretrained language models, multiple datasets, and dataset embeddings. In Proceedings of the 20th International Workshop on Treebanks and Linguistic Theories (TLT, SyntaxFest 2021), pages 96–104, Sofia, Bulgaria. Association for Computational Linguistics. ↗
  39. Rob van der Goot, Ahmet Üstün, Alan Ramponi, Ibrahim Sharaf, and Barbara Plank. 2021. Massive choice, ample tasks (MaChAmp): A toolkit for multi-task learning in NLP. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, pages 176–197, Online. Association for Computational Linguistics. 10. find our card
  40. Language Resource References Federica Gamba and Flavio Massimiliano Cecchini. 2024. De latinae linguae reparatione treebank. LINDAT/CLARIAH-CZ digital library at the Institute of Formal and Applied Linguistics (ÚFAL). find
  41. Daniel Zeman, Joakim Nivre, Mitchell Abrams, Elia Ackermann, et al. 2025. Universal dependencies 2.17. LINDAT/CLARIAH-CZ digital library at the Institute of Formal and Applied Linguistics (ÚFAL). find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/stymne-2026-uppsalanlp-evalatin-2026-multilingual-parsing/