@article{ART003353347},
author={HEUJEEUK},
title={A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules},
journal={Journal of Internet of Things and Convergence},
issn={2466-0078},
year={2026},
volume={12},
number={3},
pages={12}
TY - JOUR
AU - HEUJEEUK
TI - A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules
JO - Journal of Internet of Things and Convergence
PY - 2026
VL - 12
IS - 3
PB - The Korea Internet of Things Society
SP - 12
EP -
SN - 2466-0078
AB - This paper proposes a Transformer-based unit conversion integrated machine translation system to address the cultural unit system discrepancy that occurs during Korean-English machine translation. Existing machine translation systems transliterate units without considering the differences between the imperial and metric systems, which hinders the intuitive understanding of readers in the target cultural context. The proposed system adopts a pipeline architecture consisting of five modules: language detection, region detection, unit detection, neural machine translation, and unit conversion post-processing, utilizing a Transformer model composed of Multi-head Self-Attention and Cross-Attention as the translation engine. The target units are limited to length and temperature, and translation results are output in three modes: literal translation, unit conversion, and dual notation. Approximately 49,979 training samples were constructed from the AI Hub Korean-English parallel corpus. Experimental results demonstrate that the proposed system achieved a unit detection F1-Score of 0.91, a unit conversion average error rate of 0.11%, and a translation quality BLEU-4 score of 0.2214, confirming that the system can effectively perform unit conversion translation appropriate to the target cultural context.
KW - Machine Translation;Unit Conversion;Transformer;Cross-Attention;Imperial;Metric
DO -
UR -
ER -
HEUJEEUK. (2026). A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules. Journal of Internet of Things and Convergence, 12(3), 12.
HEUJEEUK. 2026, "A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules", Journal of Internet of Things and Convergence, vol.12, no.3 12.
HEUJEEUK "A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules" Journal of Internet of Things and Convergence 12.3 12 (2026) : 12.
HEUJEEUK. A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules. 2026; 12(3), 12.
HEUJEEUK. "A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules" Journal of Internet of Things and Convergence 12, no.3 (2026) : 12.
HEUJEEUK. A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules. Journal of Internet of Things and Convergence, 12(3), 12.
HEUJEEUK. A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules. Journal of Internet of Things and Convergence. 2026; 12(3) 12.
HEUJEEUK. A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules. 2026; 12(3), 12.
HEUJEEUK. "A Korean-English Neural Machine Translation System Utilizing Unit Detection and Conversion Modules" Journal of Internet of Things and Convergence 12, no.3 (2026) : 12.
