@article{ART003383484},
author={Rakhun Kim and Axel Deo Gratias Ndema Yetomane},
title={A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language},
journal={Journal of The Korea Society of Computer and Information},
issn={1598-849X},
year={2026},
volume={31},
number={9},
pages={31-55}
TY - JOUR
AU - Rakhun Kim
AU - Axel Deo Gratias Ndema Yetomane
TI - A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language
JO - Journal of The Korea Society of Computer and Information
PY - 2026
VL - 31
IS - 9
PB - The Korean Society Of Computer And Information
SP - 31
EP - 55
SN - 1598-849X
AB - Sango is an extremely low-resource language with limited machine translation resources. This study compares six English-to-Sango translation systems and examines the out-of-domain effects of LoRA-based adaptation. The systems were evaluated on the complete FLORES-200 devtest set using BLEU, chrF, and TER, supplemented by native-speaker judgments on a subset. Google Translate achieved the strongest overall performance, followed by zero-shot NLLB-200. All three LoRA-adapted conditions showed significantly lower BLEU scores than the zero-shot baseline in paired-bootstrap comparisons. A quantity- and step-matched comparison further showed that combining pivot-generated data, back-translation, and orthographic normalization outperformed an equal-sized condition using additional directly mined data, but did not recover zero-shot performance. Exploratory native-speaker evaluations also identified substantial semantic distortions, including unrelated religious-domain content, in some adapted outputs. Importantly, because the present experiments were conducted under a relatively small-scale adaptation setting, the observed pattern should not be assumed to generalize to substantially larger fine-tuning datasets. Overall, the results indicate that fine-tuning on automatically mined data may reduce out-of-domain performance in extremely low-resource MT, underscoring the importance of strong zero-shot baselines, independent evaluation benchmarks, and native-speaker validation.
KW - Sango;low-resource machine translation;LoRA-based fine-tuning;out-of-domain evaluation
DO -
UR -
ER -
Rakhun Kim and Axel Deo Gratias Ndema Yetomane. (2026). A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language. Journal of The Korea Society of Computer and Information, 31(9), 31-55.
Rakhun Kim and Axel Deo Gratias Ndema Yetomane. 2026, "A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language", Journal of The Korea Society of Computer and Information, vol.31, no.9 pp.31-55.
Rakhun Kim, Axel Deo Gratias Ndema Yetomane "A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language" Journal of The Korea Society of Computer and Information 31.9 pp.31-55 (2026) : 31.
Rakhun Kim, Axel Deo Gratias Ndema Yetomane. A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language. 2026; 31(9), 31-55.
Rakhun Kim and Axel Deo Gratias Ndema Yetomane. "A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language" Journal of The Korea Society of Computer and Information 31, no.9 (2026) : 31-55.
Rakhun Kim; Axel Deo Gratias Ndema Yetomane. A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language. Journal of The Korea Society of Computer and Information, 31(9), 31-55.
Rakhun Kim; Axel Deo Gratias Ndema Yetomane. A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language. Journal of The Korea Society of Computer and Information. 2026; 31(9) 31-55.
Rakhun Kim, Axel Deo Gratias Ndema Yetomane. A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language. 2026; 31(9), 31-55.
Rakhun Kim and Axel Deo Gratias Ndema Yetomane. "A Comparative Evaluation of English–Sango Machine Translation Systems: LoRA-Based Adaptation and Out-of-Domain Performance for an Extremely Low-Resource Language" Journal of The Korea Society of Computer and Information 31, no.9 (2026) : 31-55.
