AI Just Got Its First Real Shot at Translating Wolof and Arabic

There's now a hand-built, gold-standard bridge between Wolof and Modern Standard Arabic. MudawanSn is a resource of 1,271 sentence-aligned pairs manually translated from Wolof into Modern Standard Arabic. The source texts are drawn from the MasakhaNER corpus and cover politics, society, religion, and sports in Senegalese news discourse. The authors point out that although multilingual resources such as FLORES-200 and NTREX include both Wolof and Arabic, no publicly available parallel corpus is specifically designed for the Wolof–Modern Standard Arabic language pair. The paper describes the corpus construction protocol, the sentence alignment procedure, and the quality-control workflow. Four machine translation systems spanning three architectural families were benchmarked — NLLB-200 (600M), mT5-base, and two AfriNLLB variants — and fine-tuning on MudawanSn yielded substantial improvements in both translation directions. The best-performing model, AfriNLLB-12, achieves 7.76 BLEU and 30.72 chrF++ for Wolof-to-Arabic, and 8.75 BLEU and 33.08 chrF++ for Arabic-to-Wolof. The corpus is released under the CC BY-NC license and is publicly available on Hugging Face and GitHub.