Context-Aware Dialectal Arabic Machine Translation with Interactive Region and Register Selection
Afroza Nowshin, Prithweeraj Acharjee Porag, Haziq Jeelani, Fayeq Jeelani Syed · 2026
Read Context-Aware Dialectal Arabic Machine Translation with Interactive Region and Register Selection by Afroza Nowshin, Prithweeraj Acharjee Porag, Haziq Jeelani, Fayeq Jeelani Syed (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-06456.txt.
Description
Nowshin, Porag, Jeelani, and Syed build a dialectal Arabic translation system modeling regional and social variation instead of flattening dialects. Their Rule-Based Data Augmentation pipeline grows a 3,000-sentence seed into a 57,000-sentence dataset across eight varieties. An mT5-base model tuned on metadata tags controls dialect and register. NLLB scores about 14 BLEU by defaulting to the Modern Standard mean, while theirs scores about 8 BLEU but fits varieties better. Cultural authenticity rates their output 4.8 out of 5 against 1 for baselines, so metrics underrate dialect fidelity.
Catalog record
| CREATOR | Afroza Nowshin, Prithweeraj Acharjee Porag, Haziq Jeelani, Fayeq Jeelani Syed |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31176 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.06456