Elamur Research⌕
☰
home / papers / 2026 / NAJD-MT: High-Fidelity Saudi Najdi–English Training Data for Bidirectional Neural Machine Translation

NAJD-MT: High-Fidelity Saudi Najdi–English Training Data for Bidirectional Neural Machine Translation

Paper · 2026 · Nour Qandos, Samar Essa Ahmed, Omer Nacar et al.

Summary

Parallel data pairing Saudi Najdi Arabic directly with English is scarce. Pivoting through MSA can compound errors and lose dialect nuance, so the authors build NAJD-MT, a synthetic Najdi–English corpus for training translation in both directions. They take the Najdi portion of the speech-transcript dataset SADA and clean it by removing fragments, duplicates and STT artifacts. GPT-4.1, chosen after a small pilot against four other LLMs, produces the English side. Each pair is then scored by cosine similarity between Qwen3-Embedding-8B vectors of the source and its translation, and pairs below a threshold (0.6 or 0.7) are discarded. Fine-tuning NLLB-200, OPUS-MT, mBART and AraT5v2 on the two filtered versions shows that the stricter cut-off gives higher BLEU for every model and both directions, even though it removes almost half the data. Translating into English is far easier than generating Najdi. The work offers a reproducible recipe and evidence that alignment quality can matter more than corpus size for dialectal MT.

Key points

  • Corpus size: the Najdi source has 69,978 utterances, reduced to 68,158 after cleaning. About 54,280 pairs remain at τ ≥ 0.6 and 30,421 at τ ≥ 0.7, each split 80/20 into train and validation sets.
  • Filter check: manual review supported the embedding filter. Most of the 158 pairs scoring ≤ 0.3 showed serious meaning drift, while a random sample of 170 pairs at ≥ 0.7 had no critical mistranslations.
  • Najdi→English: NLLB-200 performed best, reaching 30.60 BLEU at τ = 0.7 (27.60 at τ = 0.6), with BERTScore F1 0.9387. OPUS-MT followed at 29.10 and mBART at 28.40.
  • English→Najdi: scores were much lower. mBART led at 15.60 BLEU (BERTScore F1 0.9111), just ahead of OPUS-MT at 15.10. All results are reported on validation sets.
  • Error analysis: Najdi output often mixed up gender agreement and degraded on long sentences. English output tended to repeat named entities.

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

 Institutions

🎓Tuwaiq AcademyRiyadh · Saudi Arabia13 works · 7 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 works

🧭 Topics 3Research subjects and languages the work is about.Machine translation · 1,105 worksEnglish · 2,379 worksSemitic languages · 3,066 works

Cite

Cite as Nour Qandos, Samar Essa Ahmed, Omer Nacar, Ahmad Alrabghi, Rahaf Saeed Al Hallay, Aya Hamod and Shaden Alsuhaim. 2026. NAJD-MT: High-Fidelity Saudi Najdi–English Training Data for Bidirectional Neural Machine Translation. The 7th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT7) with 5 Shared Tasks.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/27nkwba8nvda

@inproceedings{qandos2026najd,
  title = {{NAJD-MT: High-Fidelity Saudi Najdi–English Training Data for Bidirectional Neural Machine Translation}},
  author = {Nour Qandos, Samar Essa Ahmed, Omer Nacar, Ahmad Alrabghi, Rahaf Saeed {Al Hallay}, Aya Hamod and Shaden Alsuhaim},
  booktitle = {The 7th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT7) with 5 Shared Tasks},
  year = {2026},
  doi = {10.63317/27nkwba8nvda},
  url = {https://doi.org/10.63317/27nkwba8nvda}
}

References

19 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sadeen Alharbi, Areeb Alowisheq, Zoltán Tüske, Kareem Darwish, Abdullah Alrajeh, Abdulmajeed Alrowithi, Aljawharah Bin Tamran, Asma Ibrahim, Raghad Aloraini, Raneem Alnajim, et al. 2024. Sada: Saudi audio dataset for arabic. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 10286–10290. IEEE. find
  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, pages 3197–3203. ↗
  3. Houda Bouamor, Nizar Habash, Mohammad Salameh, Wajdi Zaghouani, Owen Rambow, Dana Abdulrahim, Ossama Obeid, Salam Khalifa, Fadhl Eryani, Alexander Erdmann, et al. 2018. The madar arabic dialect corpus and lexicon. In Proceedings of the eleventh international conference on language resources and evaluation (LREC 2018). find
  4. Marta R Costa-Jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, et al. 2022. No language left behind: Scaling human-centered machine translation. arXiv preprint arXiv:2207.04672. ↗
  5. 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. 2022. No language left behind: Scaling human-centered machine translation. arXiv preprint arXiv:2207.04672. ↗
  6. AbdelRahim Elmadany, El Moatez Billah Nagoudi, and Muhammad Abdul-Mageed. 2023. Octopus: A multitask model and toolkit for Arabic natural language generation. In Proceedings of Arabic-NLP 2023, pages 232–243, Singapore (Hybrid). Association for Computational Linguistics.
  7. Tianyu Gao, Xingcheng Yao, and Danqi Chen. 2021. Simcse: Simple contrastive learning of sentence embeddings. In Proceedings of the 2021 conference on empirical methods in natural language processing, pages 6894–6910.
  8. Nizar Y Habash. 2010. Introduction to Arabic natural language processing. Morgan & Claypool Publishers. ↗
  9. 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, pages 311–318. Association for Computational Linguistics.
  10. 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. ↗
  11. Wael Salloum and Nizar Habash. 2013. Dialectal arabic to english machine translation: Pivoting through modern standard arabic. In Proceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 348–358.
  12. Holger Schwenk, Guillaume Wenzek, Sergey Edunov, Edouard Grave, Armand Joulin, and Angela Fan. 2021. Ccmatrix: Mining billions of high-quality parallel sentences on the web. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 6490–6500.
  13. Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016. Improving neural machine translation models with monolingual data. In Proceedings of the 54th annual meeting of the association for computational linguistics (volume 1: long papers), pages 86–96. find
  14. Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, and Angela Fan. 2020. Multilingual translation with extensible multilingual pretraining and fine-tuning. find
  15. Jörg Tiedemann and Santhosh Thottingal. 2020. Opus-mt — building open translation services for the world. In Proceedings of the 22nd Annual Conference of the European Association for Machine Translation, pages 479–480. find
  16. Yizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu, Noah A Smith, Daniel Khashabi, and Hannaneh Hajishirzi. 2023. Self-instruct: Aligning language models with self-generated instructions. In Proceedings of the 61st annual meeting of the association for computational linguistics (volume 1: long papers), pages 13484–13508. ↗
  17. Omar F Zaidan and Chris Callison-Burch. 2014. Arabic dialect identification. Computational Linguistics, 40(1):171–202. ↗
  18. Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q. Weinberger, and Yoav Artzi. 2020. Bertscore: Evaluating text generation with bert. ↗
  19. Yanzhao Zhang, Mingxin Li, Dingkun Long, Xin Zhang, Huan Lin, Baosong Yang, Pengjun Xie, An Yang, Dayiheng Liu, Junyang Lin, Fei Huang, and Jingren Zhou. 2025. Qwen3 embedding: Advancing text embedding and reranking through foundation models. arXiv preprint arXiv:2506.05176. ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/qandos-2026-najd-mt-high-fidelity-saudi/