---
title: "Evaluating Latin and Ancient Greek — Reichbauer 2026"
description: "Finding translations and intertextual links between Ancient Greek and Latin is important for classical scholarship."
authors: ["Sebastian Reichbauer", "Shu Okabe", "Alexander Fraser"]
year: 2026
venue: "Proceedings of the 6th International Conference on Natural Language Processing for the Digital Humanities"
source: "https://aclanthology.org/2026.nlp4dh-1.11/"
doi: "10.18653/v1/2026.nlp4dh-1.11"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/reichbauer-2026-evaluating-latin-ancient-greek-sentence/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/reichbauer-2026-evaluating-latin-ancient-greek-sentence/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/6/) / Evaluating Latin and Ancient Greek Sentence Alignment through Parallel Sentence Mining

# Evaluating Latin and Ancient Greek Sentence Alignment through Parallel Sentence Mining

**Authors:** Sebastian Reichbauer, [Shu Okabe](https://md.research.elamur.ai/free/en/authors/shu-okabe/), [Alexander Fraser](https://md.research.elamur.ai/free/en/authors/alexander-fraser/) · **Year:** 2026 · **Venue:** Proceedings of the 6th International Conference on Natural Language Processing for the Digital Humanities · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.nlp4dh-1.11/) · [PDF](https://aclanthology.org/2026.nlp4dh-1.11.pdf) · [DOI 10.18653/v1/2026.nlp4dh-1.11](https://doi.org/10.18653/v1/2026.nlp4dh-1.11) · [OpenAlex](https://openalex.org/W7166810145)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Finding translations and intertextual links between Ancient Greek and Latin is important for classical scholarship. However, the two languages share very little parallel data, and existing embedding models had never been compared on this pair. The authors built a synthetic, BUCC-style mining benchmark. They inserted 2,000 Greek–Latin translation pairs, taken from Boethius' Latin versions of Aristotle's logical treatises, into monolingual corpora of roughly 22k sentences per language. They then used this benchmark to test six sentence encoders inside a CSLS-based mining pipeline. They also tried several ways to improve cross-lingual alignment:
- training-free post-processing (whitening, CBIE),
- fine-tuning on a small set of Greek–Latin parallel sentences,
- knowledge distillation with English or Latin as the anchor language.

Whitening consistently beat CBIE. Distillation through English, combined with whitening, worked best: SPhilBERTa as the student with LaBSE as the teacher reached an F-score of 97.6. The released model and benchmark give digital classicists a validated tool for retrieving Greek–Latin parallels. They are also a first step toward detecting paraphrase.

## Key points
- **Benchmark design:**
  - 2,000 parallel sentences come from Boethius' translations of *Analytica Priora* and *Topica*.
  - Distractor sentences (about 22,727 and 21,641 per side) come from authors of similar periods but cover other topics, such as historiography, theology and non-logical philosophy.
  - The Bible was excluded to avoid data contamination.
  - Preprocessing limited sentences to 10–50 words, applied NFC normalization, and standardized Latin orthography.
- **Mining pipeline:** CSLS similarity with k = 20, plus a dynamic threshold (mean + λ·σ) to decide which pairs to accept.
- **Models compared:** six encoders, including LaBSE, Ugarit, SPhilBERTa, Glot500 and Qwen3-embedding (0.6B). LaBSE, Ugarit and SPhilBERTa performed best out of the box, while Glot500 was weak.
- **Choice of anchor language:**
  - English (about 480k parallel sentences) clearly beat Latin (about 10k) as the distillation anchor.
  - With Latin as the anchor, SPhiTune was the better teacher, scoring an F-score of 90.4 after whitening, compared with 86.4 for LaBSE.
- **Release:** the best model and the benchmark are open-source at github.com/TUM-NLP/latin-greek-mining.

*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](https://md.research.elamur.ai/free/en/shelves/linguistics/) · [Natural language processing](https://md.research.elamur.ai/free/en/shelves/nlp/) · [History & humanities](https://md.research.elamur.ai/free/en/shelves/history/)

Topics: [Evaluation & benchmarks](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) · [Latin](https://md.research.elamur.ai/free/en/topics/latin/) · [Ancient Greek](https://md.research.elamur.ai/free/en/topics/ancient-greek/)

## Cite

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{reichbauer2026evaluating,
  title = {{Evaluating Latin and Ancient Greek Sentence Alignment through Parallel Sentence Mining}},
  author = {Reichbauer, Sebastian and Okabe, Shu and Fraser, Alexander},
  booktitle = {Proceedings of the 6th International Conference on Natural Language Processing for the Digital Humanities},
  year = {2026},
  doi = {10.18653/v1/2026.nlp4dh-1.11},
  url = {https://doi.org/10.18653/v1/2026.nlp4dh-1.11}
}
```

- [← previous Evaluating Humanities Theory Alignment in Large Language Models: Incremental Prompting and Statistical Assessment](https://md.research.elamur.ai/free/en/papers/pichler-2026-evaluating-humanities-theory-alignment-large/)
- [next → Evaluating LLM Style Transfer Through Readability-Based Age Assessments](https://md.research.elamur.ai/free/en/papers/maro-2026-evaluating-llm-style-transfer-through/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/6/)

## References

21 references, 3 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. Preprint, arXiv:2009.10053. [↗](https://arxiv.org/abs/2009.10053)
2. Neil Coffee, Jean-Pierre Koenig, Shakthi Poornima, Christopher W. Forstall, Roelant Ossewaarde, and Sarah L. Jacobson. 2013. The tesserae project: intertextual analysis of latin poetry. Literary and Linguistic Computing, 28(2):221–228. [↗](https://doi.org/10.1093/llc/fqs033)
3. 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. [↗](https://doi.org/10.18653/v1/2020.acl-main.747)
4. Alexis Conneau, Guillaume Lample, Marc’Aurelio Ranzato, Ludovic Denoyer, and Hervé Jégou. 2018. Word translation without parallel data. In International Conference on Learning Representations (ICLR). [↗](https://doi.org/10.48550/arxiv.1710.04087)
5. Georgiana Dinu, Angeliki Lazaridou, and Marco Baroni. 2015. Improving zero-shot learning by mitigating the hubness problem. In Proceedings of the Workshop Track at ICLR. [↗](https://doi.org/10.48550/arxiv.1412.6568)
6. Fangxiaoyu Feng, Yinfei Yang, Daniel Cer, Naveen Arivazhagan, and Wei Wang. 2022. Language-agnostic BERT sentence embedding. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 878–891, Dublin, Ireland. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Fangxiaoyu+Feng%2C+Yinfei+Yang%2C+Daniel+Cer%2C+Naveen+Arivazhagan%2C+and+Wei+Wang%2E+2022%2E+Language%2Dagnostic+BERT+sentence+embedding%2E+In+Proceedings+of+the+60th+Annual+Meeting+of+the+Association+for+Computatio) [our card](https://md.research.elamur.ai/free/en/papers/feng-2020-language-agnostic-bert-sentence-embedding/)
7. Katharina Hämmerl, Alina Fastowski, Jindřich Libovický, and Alexander Fraser. 2023. Exploring anisotropy and outliers in multilingual language models for cross-lingual semantic sentence similarity. In Findings of the Association for Computational Linguistics: ACL 2023, pages 7023–7037, Toronto, Canada. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Katharina+H%C3%A4mmerl%2C+Alina+Fastowski%2C+Jindr%CC%8Cich+Libovick%C3%BD%2C+and+Alexander+Fraser%2E+2023%2E+Exploring+anisotropy+and+outliers+in+multilingual+language+models+for+cross%2Dlingual+semantic+sentence+similarity%2E+I)
8. Viktor Hangya, Fabienne Braune, Yuliya Kalasouskaya, and Alexander Fraser. 2018. Unsupervised parallel sentence extraction from comparable corpora. In Proceedings of the 15th International Conference on Spoken Language Translation, pages 7–13, Brussels. International Conference on Spoken Language Translation. [find](https://scholar.google.com/scholar?q=Viktor+Hangya%2C+Fabienne+Braune%2C+Yuliya+Kalasouskaya%2C+and+Alexander+Fraser%2E+2018%2E+Unsupervised+parallel+sentence+extraction+from+comparable+corpora%2E+In+Proceedings+of+the+15th+International+Conference+)
9. Junjie Huang, Duyu Tang, Wanjun Zhong, Shuai Lu, Linjun Shou, Ming Gong, Daxin Jiang, and Nan Duan. 2021. WhiteningBERT: An easy unsupervised sentence embedding approach. In Findings of the Association for Computational Linguistics: EMNLP 2021, pages 238–244, Punta Cana, Dominican Republic. Association for Computational Linguistics. [↗](https://doi.org/10.60692/zzwdb-pjm18)
10. Ayyoob Imani, Peiqin Lin, Amir Hossein Kargaran, Silvia Severini, Masoud Jalili Sabet, Nora Kassner, Chunlan Ma, Helmut Schmid, André Martins, François Yvon, and Hinrich Schütze. 2023. Glot500: Scaling multilingual corpora and language models to 500 languages. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1082–1117, Toronto, Canada. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Ayyoob+Imani%2C+Peiqin+Lin%2C+Amir+Hossein+Kargaran%2C+Silvia+Severini%2C+Masoud+Jalili+Sabet%2C+Nora+Kassner%2C+Chunlan+Ma%2C+Helmut+Schmid%2C+Andr%C3%A9+Martins%2C+Fran%C3%A7ois+Yvon%2C+and+Hinrich+Sch%C3%BCtze%2E+2023%2E+Glot500%3A+Scalin)
11. Pratik Joshi, Sebastin Santy, Amar Budhiraja, Kalika Bali, and Monojit Choudhury. 2020. The state and fate of linguistic diversity and inclusion in the NLP world. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 6282–6293, Online. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Pratik+Joshi%2C+Sebastin+Santy%2C+Amar+Budhiraja%2C+Kalika+Bali%2C+and+Monojit+Choudhury%2E+2020%2E+The+state+and+fate+of+linguistic+diversity+and+inclusion+in+the+NLP+world%2E+In+Proceedings+of+the+58th+Annual+Mee) [our card](https://md.research.elamur.ai/free/en/papers/joshi-2020-state-fate-linguistic-diversity-inclusion/)
12. Kevin Krahn, Derrick Tate, and Andrew C. Lamicela. 2023. Sentence embedding models for Ancient Greek using multilingual knowledge distillation. In Proceedings of the Ancient Language Processing Workshop, pages 13–22, Varna, Bulgaria. INCOMA Ltd., Shoumen, Bulgaria. [↗](https://doi.org/10.48550/arxiv.2308.13116)
13. Shu Okabe and Alexander Fraser. 2025. Bilingual sentence mining for low-resource languages: a case study on upper and Lower Sorbian. In Proceedings of the Eight Workshop on the Use of Computational Methods in the Study of Endangered Languages, pages 11–19, Honolulu, Hawaii, USA. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Shu+Okabe+and+Alexander+Fraser%2E+2025%2E+Bilingual+sentence+mining+for+low%2Dresource+languages%3A+a+case+study+on+upper+and+Lower+Sorbian%2E+In+Proceedings+of+the+Eight+Workshop+on+the+Use+of+Computational+Me)
14. Shu Okabe, Katharina Hämmerl, and Alexander Fraser. 2025. Improving parallel sentence mining for low-resource and endangered languages. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 196–205, Vienna, Austria. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Shu+Okabe%2C+Katharina+H%C3%A4mmerl%2C+and+Alexander+Fraser%2E+2025%2E+Improving+parallel+sentence+mining+for+low%2Dresource+and+endangered+languages%2E+In+Proceedings+of+the+63rd+Annual+Meeting+of+the+Association+for)
15. Sara Rajaee and Mohammad Taher Pilehvar. 2021. A cluster-based approach for improving isotropy in contextual embedding space. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers), pages 575–584, Online. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Sara+Rajaee+and+Mohammad+Taher+Pilehvar%2E+2021%2E+A+cluster%2Dbased+approach+for+improving+isotropy+in+contextual+embedding+space%2E+In+Proceedings+of+the+59th+Annual+Meeting+of+the+Association+for+Computati)
16. Nils Reimers and Iryna Gurevych. 2020. Making monolingual sentence embeddings multilingual using knowledge distillation. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 4512–4525, Online. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Nils+Reimers+and+Iryna+Gurevych%2E+2020%2E+Making+monolingual+sentence+embeddings+multilingual+using+knowledge+distillation%2E+In+Proceedings+of+the+2020+Conference+on+Empirical+Methods+in+Natural+Language+) [our card](https://md.research.elamur.ai/free/en/papers/reimers-2020-making-monolingual-sentence-embeddings-multilingual/)
17. Frederick Riemenschneider and Anette Frank. 2023a. 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](https://scholar.google.com/scholar?q=Frederick+Riemenschneider+and+Anette+Frank%2E+2023a%2E+Exploring+large+language+models+for+classical+philology%2E+In+Proceedings+of+the+61st+Annual+Meeting+of+the+Association+for+Computational+Linguistics+%28)
18. Frederick Riemenschneider and Anette Frank. 2023b. Graecia capta ferum victorem cepit. detecting Latin allusions to Ancient Greek literature. In Proceedings of the Ancient Language Processing Workshop, pages 30–38, Varna, Bulgaria. INCOMA Ltd., Shoumen, Bulgaria. [↗](https://doi.org/10.48550/arxiv.2308.12008)
19. Jörg Tiedemann. 2012. Parallel data, tools and interfaces in opus. In Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC 2012), Istanbul, Turkey. European Language Resources Association (ELRA). [find](https://scholar.google.com/scholar?q=J%C3%B6rg+Tiedemann%2E+2012%2E+Parallel+data%2C+tools+and+interfaces+in+opus%2E+In+Proceedings+of+the+8th+International+Conference+on+Language+Resources+and+Evaluation+%28LREC+2012%29%2C+Istanbul%2C+Turkey%2E+European+Langu)
20. Tariq Yousef, Chiara Palladino, Farnoosh Shamsian, Anise d’Orange Ferreira, and Michel Ferreira dos Reis. 2022. An automatic model and gold standard for translation alignment of Ancient Greek. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 5894–5905, Marseille, France. European Language Resources Association. [find](https://scholar.google.com/scholar?q=Tariq+Yousef%2C+Chiara+Palladino%2C+Farnoosh+Shamsian%2C+Anise+d%E2%80%99Orange+Ferreira%2C+and+Michel+Ferreira+dos+Reis%2E+2022%2E+An+automatic+model+and+gold+standard+for+translation+alignment+of+Ancient+Greek%2E+In+Proc)
21. Pierre Zweigenbaum, Serge Sharoff, and Reinhard Rapp. 2017. Overview of the second BUCC shared task: Spotting parallel sentences in comparable corpora. In Proceedings of the 10th Workshop on Building and Using Comparable Corpora, pages 60–67, Vancouver, Canada. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Pierre+Zweigenbaum%2C+Serge+Sharoff%2C+and+Reinhard+Rapp%2E+2017%2E+Overview+of+the+second+BUCC+shared+task%3A+Spotting+parallel+sentences+in+comparable+corpora%2E+In+Proceedings+of+the+10th+Workshop+on+Building+)