---
title: "Rosetta at AlexandriaX-2026 — Esmaeil 2026"
description: "Rosetta is a system for Subtask 1 of the AlexandriaX 2026 shared task."
authors: ["Nada Esmaeil", "Fathima Rena", "Sibi Subhash", "Osama Elgendy", "Mina Naguib", "Salma Omar", "Muhammad Arif"]
year: 2026
venue: "arXiv:2609.10395"
source: "https://arxiv.org/abs/2609.10395"
doi: "10.48550/arxiv.2609.10395"
arxiv: "2609.10395"
licence: "CC BY 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/esmaeil-2026-rosetta-alexandriax-2026-lora-adapted/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/esmaeil-2026-rosetta-alexandriax-2026-lora-adapted/)

[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/8/) / Rosetta at AlexandriaX-2026: LoRA-Adapted NileChat for Context-Aware Dialectal Arabic Dialogue Translation

# Rosetta at AlexandriaX-2026: LoRA-Adapted NileChat for Context-Aware Dialectal Arabic Dialogue Translation

**Authors:** Nada Esmaeil, Fathima Rena, Sibi Subhash, Osama Elgendy, Mina Naguib, Salma Omar, Muhammad Arif · **Year:** 2026 · **Venue:** arXiv:2609.10395 · **Licence:** [CC BY 4.0 (arXiv)](http://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2609.10395) · [PDF](https://arxiv.org/pdf/2609.10395) · [DOI 10.48550/arxiv.2609.10395](https://doi.org/10.48550/arxiv.2609.10395) · [arXiv 2609.10395](https://arxiv.org/abs/2609.10395) · [OpenAlex](https://openalex.org/W7212178760)

**Citations:** 0 citations (openalex-snapshot · 2026-10-14) · [1](#cited-by) in our library, of the 140884 papers whose bibliographies we read

## Summary
Rosetta is a system for Subtask 1 of the AlexandriaX 2026 shared task. The task is to translate English dialogue turns into one of thirteen country-level Arabic dialects, given the earlier turns and metadata such as domain, speaker and addressee gender, and persona. Dialectal Arabic has little parallel data and no fixed spelling, so models built mostly on MSA transfer poorly. The authors freeze NileChat-3B, an Arabic instruction-tuned model, and train a LoRA adapter on it. Each training example is a system/user prompt that states the target dialect, the metadata and the running conversation. At inference, the model's own earlier translations become the history. For the unconstrained track, the adapter first goes through a pretraining stage on MADAR and PADIC, whose MSA side was machine-translated into English. The constrained system came 4th with 26.10 spBLEU. The unconstrained one came 5th with 25.09 spBLEU. So the extra data mostly hurt, which the authors frame as negative transfer they only partly understand. The result matters for anyone tempted to add loosely matched external corpora to low-resource dialect MT.

## Key points
- Setup: the base model is UBC-NLP/NileChat-3B, itself continued pretraining of Qwen2.5-3B.
  - LoRA settings: r = 32, α = 32, no dropout, applied to the q/k/v/o and gate/up/down projections.
  - Training: one epoch on a single Tesla T4 with 4-bit NF4 quantization, using Unsloth and PEFT.
- Data:
  - The Alexandria corpus has 66,480 training turns across 13 dialects, but no training turns for Libyan or Sudanese.
  - External data: 68,000 MADAR and 17,301 PADIC training pairs. Their English side was produced with facebook/nllb-200-3.3B, and MADAR city labels were collapsed to country labels.
- History noising during task fine-tuning targets exposure bias. With probability 0.15 the history is cut to a random prefix. With probability 0.25 earlier translations get word-level drops, swaps or duplications. Decoding uses beam search with 5 beams and a length penalty of 0.7.
- Track comparison: the unconstrained system was worse on 11 of 13 dialects. The exceptions were Libyan (+0.32) and Moroccan (+0.42). Sudanese dropped by 1.10 even though it also gained data where it had none before.
  - Mauritanian was by far the weakest dialect, at 12.13 spBLEU in the constrained track.
- Error analysis found two main problems:
  - Output was too literal compared with how Egyptian speakers naturally compress routine exchanges.
  - The model systematically failed to keep English code-switched words in Latin script. Missed cases: 319 Moroccan, 148 Lebanese, 123 Tunisian, 70 Egyptian. Tunisian also had the opposite error, with 212 unneeded Latin tokens.

## Interesting ideas
- During training, the context made of the model's own earlier outputs is deliberately corrupted (truncated or word-noised) while the target stays clean, so the model learns not to trust its history blindly. — For Mova (learn): when training a stage that consumes predicted upstream output (e.g. a parser fed tagger output, or frames fed parser trees), inject controlled noise into that input so the stage tolerates upstream errors.
- Instead of reporting only one average, the authors break the effect of extra pretraining down by dialect, and they let a counterexample (Sudanese) undercut the tidy explanation (Maghrebi coverage). — For Mova (learn): when new silver data or an LLM-teacher tier is added, report per-genre or per-construction deltas and look for counterexamples before accepting an explanation.
- A cheap, directional script check flags references with Latin tokens the prediction drops, and separately predictions with Latin tokens the reference lacks, turning a vague "code-switching problem" into countable error types. — For Mova (back): when generating or round-tripping text through the meta-world, count in both directions the named entities, numbers and foreign tokens that were lost or invented, as an explainable fidelity check.
- Corpora annotated at different granularity are merged by mapping fine labels onto a shared coarser scheme (cities to countries) before training. — For Mova (tree): when converting UD treebanks of different English dialects or annotation conventions, first map to a common label inventory explicitly and log the mapping, so merged training data stays traceable.
- The translation prompt carries explicit speaker/addressee gender and persona, so agreement and register come from structured facts rather than being guessed. — For Mova (back): when generating from the world of the text, pass participant attributes (gender, role, formality) from the world model directly into realization to choose pronouns, agreement and register.

*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/)

## Cited by

Cited by

1 paper of our library cites it (of the 140884 whose bibliographies we read); each has a page here.

- [AlexandriaX 2026: The First Shared Task on Dialectal Arabic Machine Translation](https://md.research.elamur.ai/free/en/papers/mekki-2026-alexandriax-2026-first-shared-task/)

## Cite

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

```bibtex
@misc{esmaeil2026rosetta,
  title = {{Rosetta at AlexandriaX-2026: LoRA-Adapted NileChat for Context-Aware Dialectal Arabic Dialogue Translation}},
  author = {Esmaeil, Nada and Rena, Fathima and Subhash, Sibi and Elgendy, Osama and Naguib, Mina and Omar, Salma and Arif, Muhammad},
  year = {2026},
  doi = {10.48550/arxiv.2609.10395},
  eprint = {2609.10395},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2609.10395}
}
```

- [← previous Romanized Arabic Across Dialects: Views, Usage Patterns, and Linguistic Variation](https://md.research.elamur.ai/free/en/papers/keleg-2026-romanized-arabic-across-dialects-views/)
- [next → Rubric-Guided Fine-tuning of SpeechLLMs for Multi-Aspect, Multi-Rater L2 Reading-Speech Assessment](https://md.research.elamur.ai/free/en/papers/parikh-2026-rubric-guided-fine-tuning-speechllms/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/8/)

## References

23 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Abdellah El Mekki, Houdaifa Atou, Omer Nacar, Shady Shehata, and Muhammad Abdul-Mageed. 2025. NileChat: Towards linguistically diverse and culturally aware LLMs for local communities. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 10967–10991, Suzhou, China. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Abdellah+El+Mekki%2C+Houdaifa+Atou%2C+Omer+Nacar%2C+Shady+Shehata%2C+and+Muhammad+Abdul%2DMageed%2E+2025%2E+NileChat%3A+Towards+linguistically+diverse+and+culturally+aware+LLMs+for+local+communities%2E+In+Proceedings+o)
2. Badr AlKhamissi, Muhammad ElNokrashy, Mai Alkhamissi, and Mona Diab. 2024. Investigating cultural alignment of large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 12404–12422, Bangkok, Thailand. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Badr+AlKhamissi%2C+Muhammad+ElNokrashy%2C+Mai+Alkhamissi%2C+and+Mona+Diab%2E+2024%2E+Investigating+cultural+alignment+of+large+language+models%2E+In+Proceedings+of+the+62nd+Annual+Meeting+of+the+Association+for+C) [our card](https://md.research.elamur.ai/free/en/papers/alkhamissi-2024-investigating-cultural-alignment-large-language/)
3. Abdellah El Mekki, AbdelRahim A. Elmadany, Samar M. Magdy, Saad Ezzini, Mo El-Haj, Mustafa Jarrar, Zaid Alyafeai, Bernard Ghanem, and Muhammad Abdul-Mageed. 2026. AlexandriaX 2026: The First Shared Task on Dialectal Arabic Machine Translation. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Abdellah+El+Mekki%2C+AbdelRahim+A%2E+Elmadany%2C+Samar+M%2E+Magdy%2C+Saad+Ezzini%2C+Mo+El%2DHaj%2C+Mustafa+Jarrar%2C+Zaid+Alyafeai%2C+Bernard+Ghanem%2C+and+Muhammad+Abdul%2DMageed%2E+2026%2E+AlexandriaX+2026%3A+The+First+Shared+Ta) [our card](https://md.research.elamur.ai/free/en/papers/mekki-2026-alexandriax-2026-first-shared-task/)
4. Yejin Bang, Samuel Cahyawijaya, Nayeon Lee, Wenliang Dai, Dan Su, Bryan Wilie, Holy Lovenia, Ziwei Ji, Tiezheng Yu, Willy Chung, Quyet V. Do, Yan Xu, and Pascale Fung. 2023. A multitask, multilingual, multimodal evaluation of ChatGPT on reasoning, hallucination, and interactivity. In Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 675–718, Nusa Dua, Bali. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Yejin+Bang%2C+Samuel+Cahyawijaya%2C+Nayeon+Lee%2C+Wenliang+Dai%2C+Dan+Su%2C+Bryan+Wilie%2C+Holy+Lovenia%2C+Ziwei+Ji%2C+Tiezheng+Yu%2C+Willy+Chung%2C+Quyet+V%2E+Do%2C+Yan+Xu%2C+and+Pascale+Fung%2E+2023%2E+A+multitask%2C+multilingual%2C) [our card](https://md.research.elamur.ai/free/en/papers/bang-2023-multitask-multilingual-multimodal-evaluation-chatgpt/)
5. Abdellah EL Mekki, Samar M. Magdy, Houdaifa Atou, Ruwa AbuHweidi, Baraah Qawasmeh, Omer Nacar, Thikra Al-hibiri, Razan Saadie, Hamzah A. Alsayadi, Nadia Ghezaiel Hammouda, Alshima Mohammed Alkhazimi, Aya Hamod, Al-Yas Yaqoob Al-Ghafri, Wesam El-Sayed, Asila Ismail al Sharji, Mohamad Ballout, Anas Belfathi, Karim Ghaddar, Serry Sibaee, and 28 others. 2026. Alexandria: A multi-domain dialectal Arabic machine translation dataset for culturally inclusive and linguistically diverse LLMs. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 32567–32592, San Diego, California, United States. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Abdellah+EL+Mekki%2C+Samar+M%2E+Magdy%2C+Houdaifa+Atou%2C+Ruwa+AbuHweidi%2C+Baraah+Qawasmeh%2C+Omer+Nacar%2C+Thikra+Al%2Dhibiri%2C+Razan+Saadie%2C+Hamzah+A%2E+Alsayadi%2C+Nadia+Ghezaiel+Hammouda%2C+Alshima+Mohammed+Alkhazimi%2C+)
6. M Saiful Bari, Yazeed Alnumay, Norah A. Alzahrani, Nouf M. Alotaibi, Hisham Abdullah Alyahya, Sultan AlRashed, Faisal Abdulrahman Mirza, Shaykhah Z. Alsubaie, Hassan A. Alahmed, Ghadah Alabduljabbar, Raghad Alkhathran, Yousef Almushayqih, Raneem Alnajim, Salman Alsubaihi, Maryam Al Mansour, Saad Amin Hassan, Dr. Majed Alrubaian, Ali Alammari, Zaki Alawami, and 7 others. 2025. ALLam: Large language models for arabic and english. In The Thirteenth International Conference on Learning Representations. Houda Bouamor, Nizar Habash, Mohammad Fanar Team, Ummar Abbas, Mohammad Shahmeer Ahmad, Firoj Alam, Enes Altinisik, Ehsannedin Asgari, Yazan Boshmaf, Sabri Boughorbel, Sanjay Chawla, Shammur Chowdhury, Fahim Dalvi, Kareem Darwish, Nadir Durrani, Mohamed Elfeky, Ahmed Elmagarmid, Mohamed Eltabakh, Masoomali Fatehkia, Anastasios Fragkopoulos, Maram Hasanain, and 23 others. 2025. Fanar: An arabic-centric multimodal generative ai platform. Preprint, arXiv:2501.13944. [↗](https://arxiv.org/abs/2501.13944) [our card](https://md.research.elamur.ai/free/en/papers/abbas-2025-fanar-arabic-centric-multimodal-generative/)
7. Karima Meftouh, Salima Harrat, Salma Jamoussi, Mourad Abbas, and Kamel Smaili. 2015. Machine translation experiments on PADIC: A parallel Arabic DIalect corpus. In Proceedings of the 29th Pacific Asia Conference on Language, Information and Computation, pages 26– 34, Shanghai, China. [find](https://scholar.google.com/scholar?q=Karima+Meftouh%2C+Salima+Harrat%2C+Salma+Jamoussi%2C+Mourad+Abbas%2C+and+Kamel+Smaili%2E+2015%2E+Machine+translation+experiments+on+PADIC%3A+A+parallel+Arabic+DIalect+corpus%2E+In+Proceedings+of+the+29th+Pacific+Asia)
8. Tarek Naous, Michael J Ryan, Alan Ritter, and Wei Xu. 2024. Having beer after prayer? measuring cultural bias in large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 16366–16393, Bangkok, Thailand. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Tarek+Naous%2C+Michael+J+Ryan%2C+Alan+Ritter%2C+and+Wei+Xu%2E+2024%2E+Having+beer+after+prayer%3F+measuring+cultural+bias+in+large+language+models%2E+In+Proceedings+of+the+62nd+Annual+Meeting+of+the+Association+for) [our card](https://md.research.elamur.ai/free/en/papers/naous-2023-having-beer-after-prayer-measuring/)
9. Charles A. Ferguson. 1959. Diglossia. WORD, 15(2):325–340. [find](https://scholar.google.com/scholar?q=Charles+A%2E+Ferguson%2E+1959%2E+Diglossia%2E+WORD%2C+15%282%29%3A325%E2%80%93340%2E)
10. Naman Goyal, Cynthia Gao, Vishrav Chaudhary, Peng-Jen Chen, Guillaume Wenzek, Da Ju, Sanjana Krishnan, Marc’Aurelio Ranzato, Francisco Guzmán, and Angela Fan. 2022. The Flores-101 evaluation benchmark for low-resource and multilingual machine translation. Transactions of the Association for Computational Linguistics, 10:522–538. [find](https://scholar.google.com/scholar?q=Naman+Goyal%2C+Cynthia+Gao%2C+Vishrav+Chaudhary%2C+Peng%2DJen+Chen%2C+Guillaume+Wenzek%2C+Da+Ju%2C+Sanjana+Krishnan%2C+Marc%E2%80%99Aurelio+Ranzato%2C+Francisco+Guzm%C3%A1n%2C+and+Angela+Fan%2E+2022%2E+The+Flores%2D101+evaluation+benchmark)
11. 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, and 20 others. 2022. No language left behind: Scaling human-centered machine translation. Preprint, arXiv:2207.04672. [↗](https://arxiv.org/abs/2207.04672)
12. Edward J Hu, yelong shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. LoRA: Low-rank adaptation of large language models. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Edward+J+Hu%2C+yelong+shen%2C+Phillip+Wallis%2C+Zeyuan+Allen%2DZhu%2C+Yuanzhi+Li%2C+Shean+Wang%2C+Lu+Wang%2C+and+Weizhu+Chen%2E+2022%2E+LoRA%3A+Low%2Drank+adaptation+of+large+language+models%2E+In+International+Conference+on+L)
13. Maja Popović. 2017. chrF++: words helping character n-grams. In Proceedings of the Second Conference on Machine Translation, pages 612–618, Copenhagen, Denmark. Association for Computational Linguistics. [↗](https://doi.org/10.18653/v1/w17-4770) [our card](https://md.research.elamur.ai/free/en/papers/popovic-2017-chrf-words-helping-character-n/)
14. Karima Kadaoui, Samar M. Magdy, Abdul Waheed, Md Tawkat Islam Khondaker, Ahmed Oumar El-Shangiti, El Moatez Billah Nagoudi, and Muhammad Abdul-Mageed. 2023. TARJAMAT: Evaluation of bard and ChatGPT on machine translation of ten Arabic varieties. In Proceedings of ArabicNLP 2023, pages 52–75, Singapore (Hybrid). Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Karima+Kadaoui%2C+Samar+M%2E+Magdy%2C+Abdul+Waheed%2C+Md+Tawkat+Islam+Khondaker%2C+Ahmed+Oumar+El%2DShangiti%2C+El+Moatez+Billah+Nagoudi%2C+and+Muhammad+Abdul%2DMageed%2E+2023%2E+TARJAMAT%3A+Evaluation+of+bard+and+ChatGPT+on) [our card](https://md.research.elamur.ai/free/en/papers/kadaoui-2023-tarjamat-evaluation-bard-chatgpt-machine/)
15. Marc’Aurelio Ranzato, Sumit Chopra, Michael Auli, and Wojciech Zaremba. 2016. Sequence level training with recurrent neural networks. Preprint, arXiv:1511.06732. [↗](https://arxiv.org/abs/1511.06732) [our card](https://md.research.elamur.ai/free/en/papers/ranzato-2015-sequence-level-training-recurrent-neural/)
16. Hassan Sajjad, Ahmed Abdelali, Nadir Durrani, and Fahim Dalvi. 2020. AraBench: Benchmarking dialectal Arabic-English machine translation. In Proceedings of the 28th International Conference on Computational Linguistics, pages 5094–5107, Barcelona, Spain (Online). International Committee on Computational Linguistics. [find](https://scholar.google.com/scholar?q=Hassan+Sajjad%2C+Ahmed+Abdelali%2C+Nadir+Durrani%2C+and+Fahim+Dalvi%2E+2020%2E+AraBench%3A+Benchmarking+dialectal+Arabic%2DEnglish+machine+translation%2E+In+Proceedings+of+the+28th+International+Conference+on+Computa)
17. Ilya Loshchilov and Frank Hutter. 2019. Decoupled weight decay regularization. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Ilya+Loshchilov+and+Frank+Hutter%2E+2019%2E+Decoupled+weight+decay+regularization%2E+In+International+Conference+on+Learning+Representations%2E)
18. Sanad Malaysha, Mo El-Haj, Saad Ezzini, Mohammed Khalilia, Mustafa Jarrar, Sultan Almujaiwel, Ismail Berrada, and Houda Bouamor. 2024. AraFinNLP 2024: The first Arabic financial NLP shared task. In Proceedings of the Second Arabic Natural Language Processing Conference, pages 393–402, Bangkok, Thailand. Association for Computational Linguistics. Neha Sengupta, Sunil Kumar Sahu, Bokang Jia, Satheesh Katipomu, Haonan Li, Fajri Koto, William Marshall, Gurpreet Gosal, Cynthia Liu, Zhiming Chen, Osama Mohammed Afzal, Samta Kamboj, Onkar Pandit, Rahul Pal, Lalit Pradhan, Zain Muhammad Mujahid, Massa Baali, Xudong Han, Sondos Mahmoud Bsharat, and 13 others. 2023. Jais and jais-chat: Arabic-centric foundation and instruction-tuned open generative large language models. Preprint, arXiv:2308.16149. [↗](https://arxiv.org/abs/2308.16149) [our card](https://md.research.elamur.ai/free/en/papers/sengupta-2023-jais-jais-chat-arabic-centric/)
19. Chihiro Taguchi, Seng Mai, Keita Kurabe, Yusuke Sakai, Georgina Agyei, Soudabeh Eslami, and David Chiang. 2025. Languages still left behind: Toward a better multilingual machine translation benchmark. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 20131–20143, Suzhou, China. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Chihiro+Taguchi%2C+Seng+Mai%2C+Keita+Kurabe%2C+Yusuke+Sakai%2C+Georgina+Agyei%2C+Soudabeh+Eslami%2C+and+David+Chiang%2E+2025%2E+Languages+still+left+behind%3A+Toward+a+better+multilingual+machine+translation+benchmark%2E)
20. Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V. Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, Jeff Klingner, Apurva Shah, Melvin Johnson, Xiaobing Liu, Łukasz Kaiser, Stephan Gouws, Yoshikiyo Kato, Taku Kudo, Hideto Kazawa, and 12 others. 2016. Google’s neural machine translation system: Bridging the gap between human and machine translation. Preprint, arXiv:1609.08144. [↗](https://arxiv.org/abs/1609.08144)
21. Jihene Younes, Emna Souissi, Achour Hadhémi, and Ahmed Ferchichi. 2020. Language resources for maghrebi arabic dialects’ nlp: a survey. Language Resources and Evaluation, 54. [find](https://scholar.google.com/scholar?q=Jihene+Younes%2C+Emna+Souissi%2C+Achour+Hadh%C3%A9mi%2C+and+Ahmed+Ferchichi%2E+2020%2E+Language+resources+for+maghrebi+arabic+dialects%E2%80%99+nlp%3A+a+survey%2E+Language+Resources+and+Evaluation%2C+54%2E)
22. Omar F. Zaidan and Chris Callison-Burch. 2014. Arabic dialect identification. Computational Linguistics, 40(1):171–202. [find](https://scholar.google.com/scholar?q=Omar+F%2E+Zaidan+and+Chris+Callison%2DBurch%2E+2014%2E+Arabic+dialect+identification%2E+Computational+Linguistics%2C+40%281%29%3A171%E2%80%93202%2E) [our card](https://md.research.elamur.ai/free/en/papers/zaidan-2014-arabic-dialect-identification/)
23. Rabih Zbib, Erika Malchiodi, Jacob Devlin, David Stallard, Spyros Matsoukas, Richard Schwartz, John Makhoul, Omar F. Zaidan, and Chris Callison-Burch. 2012. Machine translation of Arabic dialects. In Proceedings of the 2012 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 49–59, Montréal, Canada. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Rabih+Zbib%2C+Erika+Malchiodi%2C+Jacob+Devlin%2C+David+Stallard%2C+Spyros+Matsoukas%2C+Richard+Schwartz%2C+John+Makhoul%2C+Omar+F%2E+Zaidan%2C+and+Chris+Callison%2DBurch%2E+2012%2E+Machine+translation+of+Arabic+dialects%2E+In+) [our card](https://md.research.elamur.ai/free/en/papers/zbib-2012-machine-translation-arabic-dialects/)