--- title: "Lahjawi: Arabic Cross-Dialect Translator (2025)" description: "The authors introduce Lahjawi, a family of models fine-tuned from… 19.34 BLEU on Jordanian. 2 references, cited by 4 papers here." authors: ["Mohamed Motasim Hamed", "Muhammad Hreden", "Khalil Hennara", "Zeina Aldallal", "Sara Chrouf", "Safwan AlModhayan"] year: 2025 venue: "Proceedings of the 4th Workshop on Arabic Corpus Linguistics (WACL-4)" source: "https://aclanthology.org/2025.wacl-1.2/" 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/hamed-2025-lahjawi-arabic-cross-dialect-translator/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/hamed-2025-lahjawi-arabic-cross-dialect-translator/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/12/) / Lahjawi: Arabic Cross-Dialect Translator # Lahjawi: Arabic Cross-Dialect Translator **Authors:** [Mohamed Motasim Hamed](https://md.research.elamur.ai/free/en/authors/mohamed-motasim-hamed/), [Muhammad Hreden](https://md.research.elamur.ai/free/en/authors/muhammad-hreden/), [Khalil Hennara](https://md.research.elamur.ai/free/en/authors/khalil-hennara/), [Zeina Aldallal](https://md.research.elamur.ai/free/en/authors/zeina-aldallal/), [Sara Chrouf](https://md.research.elamur.ai/free/en/authors/sara-chrouf/) and [Safwan AlModhayan](https://md.research.elamur.ai/free/en/authors/safwan-almodhayan/) · **Year:** 2025 · **Venue:** [Proceedings of the 4th Workshop on Arabic Corpus Linguistics (WACL-4)](https://md.research.elamur.ai/free/en/venues/wacl/) ([Workshop on Arabic Corpus Linguistics 2025](https://md.research.elamur.ai/free/en/venues/wacl/2025/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://aclanthology.org/2025.wacl-1.2/) · [PDF](https://aclanthology.org/2025.wacl-1.2.pdf) **Citations:** [4](https://md.research.elamur.ai/free/en/papers/hamed-2025-lahjawi-arabic-cross-dialect-translator/cited-by/) in our library, of the 149,439 papers whose bibliographies we read ## Summary Arabic speakers mostly communicate in regional dialects, and these dialects differ in vocabulary, word order, idioms and morphology. Earlier machine translation work usually mapped one dialect, or a few, into Modern Standard Arabic (MSA). Direct dialect-to-dialect translation has received little attention. The authors introduce Lahjawi, a family of models fine-tuned from Kuwain-1.5B, their own small Arabic language model. - **Lahjawi-D2D** translates directly among 15 dialects. - **Lahjawi-D2MSA** normalizes any dialect into MSA. The training data comes from seven open parallel corpora, cast into two prompt templates. For the dialect-to-dialect set, every possible pairing of aligned dialect sentences was generated. Evaluation used NADI-2024 and the MADAR test set, plus human judgments. BLEU is modest: about 9.6 for D2MSA and 9.9 for D2D. Human raters judged 58% of outputs accurate and 78% fluent. Results are strongest for Levantine dialects and weakest for Maghrebi ones. The paper matters mainly as the first broad cross-dialect Arabic translator and as a map of where data imbalance and dialect distance hurt. ## Key points - **Training data.** Sources are MADAR, PADIC, NADI (2023), Dial2MSA, Arabic STS, UFAL Parallel Corpus of North Levantine 1.0 and MDPCA, normalized for characters, numerals, punctuation and spacing. - **D2MSA set.** It has 197,042 samples and is heavily skewed, with Syrian making up 66% and other dialects between 0.8% and 5.7%. - **D2D set.** It has 266,871 samples spanning 210 directed pairs over 15 dialects, and Levantine and Maghrebi pairs are over-represented. - **Evaluation and comparison.** Testing used NADI-2024 DA-MSA (Egyptian, Emirati, Jordanian, Palestinian) and the MADAR test set. Lahjawi-D2MSA trails shared-task systems built on Jais-13B or LLaMA-8B, or on augmented data. - **Findings.** - A Palestinian-trained model scored 19.34 BLEU on Jordanian data, while the Jordanian-trained model scored only 5.67 on its own dialect. - Translation is direction-dependent: Qatari→Iraqi reached 17.21 BLEU, but Iraqi→Qatari only 6.02. ## Interesting ideas - Building all ordered pairs from sentences aligned through a shared pivot turns a multi-way parallel corpus into training data for every direction at once. - A single model serves many directions, with the target variety named explicitly inside a fixed prompt template, instead of separate models per pair. - Cross-testing models across neighbouring varieties showed that a small dataset, not dialect difficulty, explained a weak result, because pooling similar varieties helped. - Scores were reported per direction and revealed strong asymmetry, so averaging both directions would hide real failures. - Weak results were explained by checking which varieties the base model likely saw during pre-training, not only fine-tuning data counts. ## For Mova How these ideas could serve Mova, our small open language model — its files, stages and tests: - **I1** (back): Sentences aligned across several English registers or paraphrase sources could be expanded into all-direction pairs to test text → graph → text style transfer through the meta-world. - **I2** (learn): LLM teacher tiers could use one templated prompt with an explicit target-schema slot (e.g., UD version or frame inventory), keeping silver data uniform and comparable. - **I3** (tree): When gparse underperforms on a small English treebank, train on one treebank and test on another, then pool related treebanks, to separate data scarcity from genuine domain mismatch. - **I4** (tree): UD dialect conversion rules should be evaluated A→B and B→A separately, with round-trip checks, to expose lossy conversion directions. - **I5** (learn): Evaluation reports could log the base model's pre-training exposure per genre or source, so gains are not wrongly credited to Mova's own training. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* 📚 Shelves (2): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works 🧭 Topics (1): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 2,908 works ## Cite Cite as: Mohamed Motasim Hamed, Muhammad Hreden, Khalil Hennara, Zeina Aldallal, Sara Chrouf and Safwan AlModhayan. 2025. Lahjawi: Arabic Cross-Dialect Translator. Proceedings of the 4th Workshop on Arabic Corpus Linguistics (WACL-4). The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{hamed2025lahjawi, title = {{Lahjawi: Arabic Cross-Dialect Translator}}, author = {Mohamed Motasim Hamed, Muhammad Hreden, Khalil Hennara, Zeina Aldallal, Sara Chrouf and Safwan AlModhayan}, booktitle = {Proceedings of the 4th Workshop on Arabic Corpus Linguistics (WACL-4)}, year = {2025}, url = {https://aclanthology.org/2025.wacl-1.2/} } ``` - [← previous Kuwain 1.5B: An Arabic SLM via Language Injection](https://md.research.elamur.ai/free/en/papers/hennara-2025-kuwain-1-5b-arabic-slm/) - [next → Large Language Models as a Normalizer for Transliteration and Dialectal Translation](https://md.research.elamur.ai/free/en/papers/alam-2025-large-language-models-as-normalizer/) - [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/12/) ## References 2 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. AhmedElmogtaba Abdelmoniem Ali Abdelaziz, Ashraf Hatim Elneima, and Kareem Darwish. 2024. LLM-based MT data creation: Dialectal to MSA translation shared task. In Proceedings of the 6th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT) with Shared Tasks on Arabic LLMs Hallucination and Dialect to MSA Machine Translation @ LREC-COLING 2024, pages 112--116, Torino, Italia. ELRA and ICCL. [find](https://scholar.google.com/scholar?q=AhmedElmogtaba+Abdelmoniem+Ali+Abdelaziz%2C+Ashraf+Hatim+Elneima%2C+and+Kareem+Darwish%2E+2024%2E+LLM%2Dbased+MT+data+creation%3A+Dialectal+to+MSA+translation+shared+task%2E+In+Proceedings+of+the+6th+Workshop+on+Op) 2. Muhammad Abdul-Mageed, AbdelRahim Elmadany, Chiyu Zhang, El Moatez Billah Nagoudi, Houda Bouamor, and Nizar Habash. 2023. [NADI 2023: The fourth nuanced Arabic dialect identification shared task](https://md.research.elamur.ai/free/en/papers/mageed-2023-nadi-2023-fourth-nuanced-arabic/). In Proceedings of ArabicNLP 2023, pages 600--613, Singapore (Hybrid). Association for Computational Linguistics. *Updated 31 October 2026*