---
title: "ParaCLEAN: Improving Translation Quality through — Mash 2026"
description: "Web-collected parallel corpora are noisy, especially in low-resource settings that mix authentic and synthetic pairs."
authors: ["Audrey Mash", "Ella Paulina Bohman", "Maite Melero"]
year: 2026
venue: "Proceedings of the Fifteenth Language Resources and Evaluation Conference"
source: "https://aclanthology.org/2026.lrec-1.527/"
doi: "10.63317/36e3vfurjna4"
licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)"
licence_url: "https://creativecommons.org/licenses/by-nc/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/papers/mash-2026-paraclean-improving-translation-quality-through/"
---
[Elamur Research](https://md.research.elamur.ai/) · [this page in HTML](https://research.elamur.ai/free/papers/mash-2026-paraclean-improving-translation-quality-through/)

[home](https://md.research.elamur.ai/) / [papers](https://md.research.elamur.ai/free/papers/) / [2026](https://md.research.elamur.ai/free/papers/2026/page/4/) / ParaCLEAN: Improving Translation Quality through Systematic Parallel Data Cleaning

# ParaCLEAN: Improving Translation Quality through Systematic Parallel Data Cleaning

**Authors:** Audrey Mash, Ella Paulina Bohman, [Maite Melero](https://md.research.elamur.ai/free/authors/maite-melero/) · **Year:** 2026 · **Venue:** Proceedings of the Fifteenth Language Resources and Evaluation Conference · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.lrec-1.527/) · [PDF](https://aclanthology.org/2026.lrec-1.527.pdf) · [DOI 10.63317/36e3vfurjna4](https://doi.org/10.63317/36e3vfurjna4) · [OpenAlex](https://openalex.org/W7160071024)

**Citations:** 0 citations (openalex-snapshot · 2026-10-14)

## Summary
Web-collected parallel corpora are noisy, especially in low-resource settings that mix authentic and synthetic pairs. Typical problems are misalignments, wrong-language segments, repeated content and inconsistent formatting. Mash, Bohman and Melero address this with ParaCLEAN, an open-source command-line pipeline. Its stages can be enabled, disabled or reordered through a configuration file:

* scoring each pair with cross-lingual sentence embeddings
* checking both sides' expected language with GlotLID
* discarding pairs below user-set thresholds
* removing exact and near duplicates
* optionally repairing text with Bifixer
* applying language-specific normalisation

Supervised filters such as Bicleaner need clean training data for each pair. ParaCLEAN does not, so it also works for pairs without English. The authors test it on Catalan→Japanese MT, using OPUS corpora plus Spanish–Japanese data whose Spanish side was machine-translated into Catalan. They report gains in both data quality and translation performance, and ablations isolate each step's contribution. The provided excerpt omits the exact metric scores. The work offers a lightweight, reproducible, language-agnostic cleaning baseline.

## Key points
- **Embeddings:** LaBSE is the default model, downloaded automatically. SONAR can be used through a local install. Cosine similarity for each pair is stored as metadata for threshold filtering.
- **Language identification:** The authors compared Lingua-py, fastText, Idiomata Cognitor and GlotLID on Flores+ and WMT24++. GlotLID scored best, with 91.07% accuracy on Flores+ and 87.92% on the noisier WMT24++.
- **Deduplication:** Text is first heavily normalised (lowercased, non-letters and accents stripped, whitespace collapsed). The source and target are then hashed together and passed through an external stable sort. Among colliding pairs, the most content-rich one is kept.
- **Data:** 1,698,452 authentic Catalan–Japanese pairs come from KDE4, MultiCCAligned, OpenSubtitles, WikiMatrix and XLEnt. 24,601,795 synthetic pairs come from Spanish–Japanese corpora (MultiCCAligned, NLLB, OpenSubtitles, WikiMatrix), with Spanish translated into Catalan by SalamandraTA-7b.
- **Release:** Code and configs are at github.com/langtech-bsc/ParaCLEAN. Bifixer is an optional separate install, and the authors recommend running it with its own normalisation and deduplication turned off.

*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/shelves/linguistics/) · [Machine translation](https://md.research.elamur.ai/free/shelves/machine-translation/)

## Cite

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

```bibtex
@inproceedings{mash2026paraclean,
  title = {{ParaCLEAN: Improving Translation Quality through Systematic Parallel Data Cleaning}},
  author = {Mash, Audrey and Bohman, Ella Paulina and Melero, Maite},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/36e3vfurjna4},
  url = {https://doi.org/10.63317/36e3vfurjna4}
}
```

- [← previous OpenSubtitles2024: A Massively Parallel Dataset of Movie Subtitles for MT Development and Evaluation](https://md.research.elamur.ai/free/papers/tiedemann-2026-opensubtitles2024-massively-parallel-dataset-movie/)
- [next → Paragraph Segmentation Revisited: Towards a Standard Task for Structuring Speech](https://md.research.elamur.ai/free/papers/retkowski-2026-paragraph-segmentation-revisited-standard-task/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/papers/2026/page/4/)

## References

32 references, 15 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Haluk Açarçiçek, Talha Çolakoğlu, Pınar Ece Aktan Hatipoğlu, Chong Hsuan Huang, and Wei Peng. 2020. Filtering noisy parallel corpus using transformers with proxy task learning. In Proceedings of the Fifth Conference on Machine Translation, pages 940–946, Online. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Haluk+A%C3%A7ar%C3%A7i%C3%A7ek%2C+Talha+%C3%87olako%C4%9Flu%2C+P%C4%B1nar+Ece+Aktan+Hatipo%C4%9Flu%2C+Chong+Hsuan+Huang%2C+and+Wei+Peng%2E+2020%2E+Filtering+noisy+parallel+corpus+using+transformers+with+proxy+task+learning%2E+In+Proceedings+of+the+F)
2. Mikel Artetxe and Holger Schwenk. 2019. Margin-based parallel corpus mining with multilingual sentence embeddings. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, page 3197–3203. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/p19-1309) [our card](https://md.research.elamur.ai/free/papers/artetxe-2019-margin-based-parallel-corpus-mining/)
3. Mikko Aulamo, Sami Virpioja, and Jörg Tiedemann. 2020. OpusFilter: A configurable parallel corpus filtering toolbox. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 150–156, Online. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Mikko+Aulamo%2C+Sami+Virpioja%2C+and+J%C3%B6rg+Tiedemann%2E+2020%2E+OpusFilter%3A+A+configurable+parallel+corpus+filtering+toolbox%2E+In+Proceedings+of+the+58th+Annual+Meeting+of+the+Association+for+Computational+Ling)
4. Daniel Deutsch, Eleftheria Briakou, Isaac Caswell, Mara Finkelstein, Rebecca Galor, Juraj Juraska, Geza Kovacs, Alison Lui, Ricardo Rei, Jason Riesa, Shruti Rijhwani, Parker Riley, Elizabeth Salesky, Firas Trabelsi, Stephanie Winkler, Biao Zhang, and Markus Freitag. 2025. Wmt24++: Expanding the language coverage of wmt24 to 55 languages dialects. [↗](https://doi.org/10.48550/arxiv.2502.12404)
5. Paul-Ambroise Duquenne, Holger Schwenk, and Benoît Sagot. 2023. Sonar: Sentence-level multimodal and language-agnostic representations. [↗](https://doi.org/10.48550/arxiv.2308.11466)
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/papers/feng-2020-language-agnostic-bert-sentence-embedding/)
7. Aarón Galiano-Jiménez, Felipe Sánchez-Martínez, and Juan Antonio Pérez-Ortiz. 2024. Idiomata cognitor. [find](https://scholar.google.com/scholar?q=Aar%C3%B3n+Galiano%2DJim%C3%A9nez%2C+Felipe+S%C3%A1nchez%2DMart%C3%ADnez%2C+and+Juan+Antonio+P%C3%A9rez%2DOrtiz%2E+2024%2E+Idiomata+cognitor%2E)
8. Javier Garcia Gilabert, Xixian Liao, Severino Da Dalt, Ella Bohman, Audrey Mash, Francesca De Luca Fornaciari, Irene Baucells, Joan Llop, Miguel Claramunt Argote, Carlos Escolano, and Maite Melero. 2025. From salamandra to salamandrata: Bsc submission for wmt25 general machine translation shared task. [find](https://scholar.google.com/scholar?q=Javier+Garcia+Gilabert%2C+Xixian+Liao%2C+Severino+Da+Dalt%2C+Ella+Bohman%2C+Audrey+Mash%2C+Francesca+De+Luca+Fornaciari%2C+Irene+Baucells%2C+Joan+Llop%2C+Miguel+Claramunt+Argote%2C+Carlos+Escolano%2C+and+Maite+Melero%2E+20)
9. Armand Joulin, Edouard Grave, Piotr Bojanowski, and Tomas Mikolov. 2017. Bag of tricks for efficient text classification. In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2017), pages 427–431. [find](https://scholar.google.com/scholar?q=Armand+Joulin%2C+Edouard+Grave%2C+Piotr+Bojanowski%2C+and+Tomas+Mikolov%2E+2017%2E+Bag+of+tricks+for+efficient+text+classification%2E+In+Proceedings+of+the+15th+Conference+of+the+European+Chapter+of+the+Associati)
10. Amir Hossein Kargaran, Ayyoob Imani, François Yvon, and Hinrich Schütze. 2023. GlotLID: Language identification for low-resource languages. In The 2023 Conference on Empirical Methods in Natural Language Processing. [find](https://scholar.google.com/scholar?q=Amir+Hossein+Kargaran%2C+Ayyoob+Imani%2C+Fran%C3%A7ois+Yvon%2C+and+Hinrich+Sch%C3%BCtze%2E+2023%2E+GlotLID%3A+Language+identification+for+low%2Dresource+languages%2E+In+The+2023+Conference+on+Empirical+Methods+in+Natural+Langu) [our card](https://md.research.elamur.ai/free/papers/kargaran-2023-glotlid-language-identification-low-resource/)
11. Huda Khayrallah and Philipp Koehn. 2018. On the impact of various types of noise on neural machine translation. arXiv preprint arXiv:1805.12282. [↗](https://arxiv.org/abs/1805.12282) [our card](https://md.research.elamur.ai/free/papers/khayrallah-2018-impact-various-types-noise-neural/)
12. Diederik P. Kingma and Jimmy Ba. 2015. Adam: A method for stochastic optimization. In ICLR (Poster). [↗](https://doi.org/10.48550/arxiv.1412.6980)
13. Philipp Koehn, Vishrav Chaudhary, Ahmed El-Kishky, Naman Goyal, Peng-Jen Chen, and Francisco Guzmán. 2020. Findings of the wmt 2020 shared task on parallel corpus filtering and alignment. In Proceedings of the Fifth Conference on Machine Translation, pages 726–742. [find](https://scholar.google.com/scholar?q=Philipp+Koehn%2C+Vishrav+Chaudhary%2C+Ahmed+El%2DKishky%2C+Naman+Goyal%2C+Peng%2DJen+Chen%2C+and+Francisco+Guzm%C3%A1n%2E+2020%2E+Findings+of+the+wmt+2020+shared+task+on+parallel+corpus+filtering+and+alignment%2E+In+Proceedin) [our card](https://md.research.elamur.ai/free/papers/koehn-2020-findings-wmt-2020-shared-task/)
14. Philipp Koehn, Francisco Guzmán, Vishrav Chaudhary, and Juan Pino. 2019. Findings of the wmt 2019 shared task on parallel corpus filtering for low-resource conditions. In Proceedings of the Fourth Conference on Machine Translation (Volume 3: Shared Task Papers, Day 2), pages 54– 72. [find](https://scholar.google.com/scholar?q=Philipp+Koehn%2C+Francisco+Guzm%C3%A1n%2C+Vishrav+Chaudhary%2C+and+Juan+Pino%2E+2019%2E+Findings+of+the+wmt+2019+shared+task+on+parallel+corpus+filtering+for+low%2Dresource+conditions%2E+In+Proceedings+of+the+Fourth+Con) [our card](https://md.research.elamur.ai/free/papers/koehn-2019-findings-wmt-2019-shared-task/)
15. Philipp Koehn, Huda Khayrallah, Kenneth Heafield, and Mikel L Forcada. 2018. Findings of the wmt 2018 shared task on parallel corpus filtering. In EMNLP 2018 Third Conference on Machine Translation (WMT18), pages 726–739. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Philipp+Koehn%2C+Huda+Khayrallah%2C+Kenneth+Heafield%2C+and+Mikel+L+Forcada%2E+2018%2E+Findings+of+the+wmt+2018+shared+task+on+parallel+corpus+filtering%2E+In+EMNLP+2018+Third+Conference+on+Machine+Translation+%28W) [our card](https://md.research.elamur.ai/free/papers/koehn-2018-findings-wmt-2018-shared-task/)
16. Julia Kreutzer, Isaac Caswell, Lisa Wang, Ahsan Wahab, Daan Van Esch, Nasanbayar Ulzii-Orshikh, Allahsera Tapo, Nishant Subramani, Artem Sokolov, Claytone Sikasote, et al. 2022. Quality at a glance: An audit of web-crawled multilingual datasets. Transactions of the Association for Computational Linguistics, 10:50–72. [↗](https://doi.org/10.1162/tacl_a_00447) [our card](https://md.research.elamur.ai/free/papers/kreutzer-2021-quality-glance-audit-web-crawled/)
17. Taku Kudo and John Richardson. 2018. Sentencepiece: A simple and language independent subword tokenizer and detokenizer for neural text processing. CoRR, abs/1808.06226. [↗](https://arxiv.org/abs/1808.06226) [our card](https://md.research.elamur.ai/free/papers/kudo-2018-sentencepiece-simple-language-independent-subword/)
18. Katherine Lee, Daphne Ippolito, Andrew Nystrom, Chiyuan Zhang, Douglas Eck, Chris Callison-Burch, and Nicholas Carlini. 2022. Deduplicating training data makes language models better. [↗](https://doi.org/10.48550/arxiv.2107.06499)
19. Jun Lu, Xin Ge, Yangbin Shi, and Yuqi Zhang. 2020. Alibaba submission to the wmt20 parallel corpus filtering task. In Proceedings of the Fifth Conference on Machine Translation, pages 979–984, Online. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Jun+Lu%2C+Xin+Ge%2C+Yangbin+Shi%2C+and+Yuqi+Zhang%2E+2020%2E+Alibaba+submission+to+the+wmt20+parallel+corpus+filtering+task%2E+In+Proceedings+of+the+Fifth+Conference+on+Machine+Translation%2C+pages+979%E2%80%93984%2C+Online%2E)
20. Alexandra Sasha Luccioni and Joseph D Viviano. 2021. What’s in the box? a preliminary analysis of undesirable content in the common crawl corpus. arXiv preprint arXiv:2105.02732. [↗](https://arxiv.org/abs/2105.02732)
21. NLLB Team, Marta R. Costa-jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, Anna Sun, Skyler Wang, Guillaume Wenzek, Al Youngblood, Bapi Akula, Loic Barrault, Gabriel Mejia Gonzalez, Prangthip Hansanti, John Hoffman, Semarley Jarrett, Kaushik Ram Sadagopan, Dirk Rowe, Shannon Spruit, Chau Tran, Pierre Andrews, Necip Fazil Ayan, Shruti Bhosale, Sergey Edunov, Angela Fan, Cynthia Gao, Vedanuj Goswami, Francisco Guzmán, Philipp Koehn, Alexandre Mourachko, Christophe Ropers, Safiyyah Saleem, Holger Schwenk, and Jeff Wang. 2024. Scaling neural machine translation to 200 languages. Nature, 630(8018):841–846. [↗](https://doi.org/10.1038/s41586-024-07335-x) [our card](https://md.research.elamur.ai/free/papers/ott-2018-scaling-neural-machine-translation/)
22. Myle Ott, Sergey Edunov, Alexei Baevski, Angela Fan, Sam Gross, Nathan Ng, David Grangier, and Michael Auli. 2019. fairseq: A fast, extensible toolkit for sequence modeling. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics (Demonstrations), pages 48–53, Minneapolis, Minnesota. Association for Computational Linguistics. [↗](https://doi.org/10.48550/arxiv.1904.01038) [our card](https://md.research.elamur.ai/free/papers/ott-2019-fairseq-fast-extensible-toolkit-sequence/)
23. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics (ACL), pages 311–318. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Kishore+Papineni%2C+Salim+Roukos%2C+Todd+Ward%2C+and+Wei%2DJing+Zhu%2E+2002%2E+Bleu%3A+a+method+for+automatic+evaluation+of+machine+translation%2E+In+Proceedings+of+the+40th+Annual+Meeting+of+the+Association+for+Comp) [our card](https://md.research.elamur.ai/free/papers/papineni-2002-bleu-method-automatic-evaluation-machine/)
24. Maja Popović. 2015. chrF: character n-gram Fscore for automatic MT evaluation. In Proceedings of the Tenth Workshop on Statistical Machine Translation (WMT), pages 392–395. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Maja+Popovi%C4%87%2E+2015%2E+chrF%3A+character+n%2Dgram+Fscore+for+automatic+MT+evaluation%2E+In+Proceedings+of+the+Tenth+Workshop+on+Statistical+Machine+Translation+%28WMT%29%2C+pages+392%E2%80%93395%2E+Association+for+Computation)
25. Gema Ramírez-Sánchez, Jaume Zaragoza-Bernabeu, Marta Bañón, and Sergio Ortiz-Rojas. 2020. Bifixer and bicleaner: two open-source tools to clean your parallel data. In Proceedings of the 22nd Annual Conference of the European Association for Machine Translation, pages 291– 298, Lisboa, Portugal. European Association for Machine Translation. [find](https://scholar.google.com/scholar?q=Gema+Ram%C3%ADrez%2DS%C3%A1nchez%2C+Jaume+Zaragoza%2DBernabeu%2C+Marta+Ba%C3%B1%C3%B3n%2C+and+Sergio+Ortiz%2DRojas%2E+2020%2E+Bifixer+and+bicleaner%3A+two+open%2Dsource+tools+to+clean+your+parallel+data%2E+In+Proceedings+of+the+22nd+Annual+Co) [our card](https://md.research.elamur.ai/free/papers/sanchez-2020-bifixer-bicleaner-two-open-source/)
26. Ricardo Rei, Ana Farinha, Alon Lavie, and André F. T. Martins. 2020. COMET: A neural framework for MT evaluation. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 2685–2702. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Ricardo+Rei%2C+Ana+Farinha%2C+Alon+Lavie%2C+and+Andr%C3%A9+F%2E+T%2E+Martins%2E+2020%2E+COMET%3A+A+neural+framework+for+MT+evaluation%2E+In+Proceedings+of+the+2020+Conference+on+Empirical+Methods+in+Natural+Language+Process) [our card](https://md.research.elamur.ai/free/papers/rei-2020-comet-neural-framework-mt-evaluation/)
27. Steve Sloto, Brian Thompson, Huda Khayrallah, Tobias Domhan, Thamme Gowda, and Philipp Koehn. 2023. Findings of the WMT 2023 shared task on parallel data curation. In Proceedings of the Eighth Conference on Machine Translation, pages 95–102, Singapore. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Steve+Sloto%2C+Brian+Thompson%2C+Huda+Khayrallah%2C+Tobias+Domhan%2C+Thamme+Gowda%2C+and+Philipp+Koehn%2E+2023%2E+Findings+of+the+WMT+2023+shared+task+on+parallel+data+curation%2E+In+Proceedings+of+the+Eighth+Confere)
28. Matthew Snover, Bonnie Dorr, Richard Schwartz, Linnea Micciulla, and John Makhoul. 2006. A study of translation edit rate with targeted human annotation. In Proceedings of the 7th Conference of the Association for Machine Translation in the Americas (AMTA), pages 223–231. Association for Machine Translation in the Americas. [find](https://scholar.google.com/scholar?q=Matthew+Snover%2C+Bonnie+Dorr%2C+Richard+Schwartz%2C+Linnea+Micciulla%2C+and+John+Makhoul%2E+2006%2E+A+study+of+translation+edit+rate+with+targeted+human+annotation%2E+In+Proceedings+of+the+7th+Conference+of+the+As) [our card](https://md.research.elamur.ai/free/papers/snover-2006-study-translation-edit-rate-targeted/)
29. Jörg Tiedemann. 2012. Parallel data, tools and interfaces in OPUS. In Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC’12), pages 2214–2218, 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+Eighth+International+Conference+on+Language+Resources+and+Evaluation+%28LREC%E2%80%9912%29%2C+pages+2214%E2%80%932218%2C+Istanbul%2C+Turk)
30. Rik Van Noord, Miquel Esplà-Gomis, Mălina Chichirău, Gema Ramírez-Sánchez, and Antonio Toral. 2025. Quality beyond a glance: Revealing large quality differences between web-crawled parallel corpora. In Proceedings of the 31st International Conference on Computational Linguistics, pages 1824–1838. [find](https://scholar.google.com/scholar?q=Rik+Van+Noord%2C+Miquel+Espl%C3%A0%2DGomis%2C+M%C4%83lina+Chichir%C4%83u%2C+Gema+Ram%C3%ADrez%2DS%C3%A1nchez%2C+and+Antonio+Toral%2E+2025%2E+Quality+beyond+a+glance%3A+Revealing+large+quality+differences+between+web%2Dcrawled+parallel+corpora%2E+I)
31. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. CoRR, abs/1706.03762. [↗](https://arxiv.org/abs/1706.03762)
32. Jaume Zaragoza-Bernabeu, Gema Ramírez-Sánchez, Marta Bañón, and Sergio Ortiz Rojas. 2022. Bicleaner ai: Bicleaner goes neural. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 824–831. 10. Appendix A: Parallel Data Sources Dataset Licence Sentence pairs Catalan–Japanese (authentic) KDE4 (v2) MultiCCAligned (v1.1) OpenSubtitles (v2024) WikiMatrix (v1) XLEnt (v1.2) CC BY 4.0 CC BY 4.0 † CC BY 4.0 CC BY-SA 4.0 CC BY 4.0 Total Spanish–Japanese (synthetic, es→ca) MultiCCAligned (v1.1) NLLB (v1) OpenSubtitles (v2024) WikiMatrix (v1) CC BY 4.0 † ODC-BY CC BY 4.0 CC BY-SA 4.0 7,057,829 15,402,228 1,922,477 219,261 Total 24,601,795 Table 8: Parallel data sources and sizes for Catalan–Japanese and Spanish–Japanese datasets. All corpora were obtained from OPUS. The Spanish side of synthetic corpora was translated into Catalan using SalamandraTA-7b. terms of agreement. 120,969 804,536 116,270 103,899 552,778 1,698,452 [find](https://scholar.google.com/scholar?q=Jaume+Zaragoza%2DBernabeu%2C+Gema+Ram%C3%ADrez%2DS%C3%A1nchez%2C+Marta+Ba%C3%B1%C3%B3n%2C+and+Sergio+Ortiz+Rojas%2E+2022%2E+Bicleaner+ai%3A+Bicleaner+goes+neural%2E+In+Proceedings+of+the+Thirteenth+Language+Resources+and+Evaluation+Confer)