Abstract
AbstractAdvances in neural machine translation utilizing pretrained language models (PLMs) have shown promise in improving the translation quality between diverse languages. However, translation from English to languages with complex morphology, such as Arabic, remains challenging. This study investigated the prevailing error patterns of state-of-the-art PLMs when translating from English to Arabic across different text domains. Through empirical analysis using automatic metrics (chrF, BERTScore, COMET) and manual evaluation with the Multidimensional Quality Metrics (MQM) framework, we compared Google Translate and five PLMs (Helsinki, Marefa, Facebook, GPT-3.5-turbo, and GPT-4). Key findings provide valuable insights into current PLM limitations in handling aspects of Arabic grammar and vocabulary while also informing future improvements for advancing English–Arabic machine translation capabilities and accessibility.
Publisher
Springer Science and Business Media LLC
Reference39 articles.
1. Zaugg IA, Hossain A, Molloy B. Digitally-disadvantaged languages. Internet Policy Rev J Internet Regul. 2022;11(2):1–11.
2. Patil A, Joshi I, Kadam D. PICT@WAT 2022: neural machine translation systems for indic languages. In: Proceedings of the 9th workshop on Asian Translation, Gyeongju, Republic of Korea: international conference on computational linguistics. 2022. pp. 106–110. https://aclanthology.org/2022.wat-1.13. Accessed 20 Dec 2023.
3. Chen K, Wang R, Utiyama M, Sumita E. Integrating prior translation knowledge into neural machine translation. IEEEACM Trans Audio Speech Lang Process. 2022;30:330–9. https://doi.org/10.1109/TASLP.2021.3138714.
4. Akan MF, Karim MR, Chowdhury AMK. An analysis of Arabic–English translation: problems and prospects. Adv Lang Lit Stud. 2019;10(1):58–65. https://doi.org/10.7575/aiac.alls.v.10n.1p.58.
5. Mamoori MMA, Tarish AH, Hasani SA. Difficulties of translation and evaluative idioms in English and Arabic. Int J Health Sci. 2022. https://doi.org/10.53730/ijhs.v6nS5.10039.