A Comparison of Human and DeepL Translations Based on Quantitative Linguistic Features
Arab World English Journal for Translation and Literary Studies · 20.10.2026 · 10 (4) · 48-60
DOI: 10.24093/awejtls/vol10no4.4 ↗Kurz gefasst
Fachartikel in der Zeitschrift Arab World English Journal for Translation and Literary Studies, online erschienen am 20.10.2026. Verfasst von Xingwen Peng und 1 weiteren Autor:innen.
Abstract (Originalsprache)
In the era of artificial intelligence, the rapid advancement of translation technology has greatly promoted the development of the translation industry. This study aims to, by using the human translations and DeepL translations translated from the same original translation competition texts as the research objects, compare the human translations and DeepL translations by focusing on their quantitative linguistic characteristics. The questions guiding the research include what the differences are between DeepL translations and human translations in English-Chinese translation in terms of quantitative linguistic characteristics and how DeepL can improve its English-Chinese translation. This study employs corpus methods to analyze nine linguistic features in human and DeepL translations, and compare the differences in these features between the human and DeepL translations. The study results indicate that human translations show greater performance in lexical richness and density, average word length, idiom use, adverb use and hapax legomenon, which is associated with a more formal and concise style. DeepL translations show more lexical repetition, a higher proportion of conjunctions and particles, fewer idioms, and longer average sentence length, which suggests a narrower range of vocabulary choices and a tendency to retain source-language syntactic structures to make longer sentences. The significance of this study is that it provides evidence to optimize DeepL translation, as a kind of neural machine translation system, in English-Chinese translation.
Zitieren
APA
Peng, X., & Duan, H. (2026). A Comparison of Human and DeepL Translations Based on Quantitative Linguistic Features. Arab World English Journal for Translation and Literary Studies, 10(4), 48-60. https://doi.org/10.24093/awejtls/vol10no4.4
Vancouver
Peng X, Duan H. A Comparison of Human and DeepL Translations Based on Quantitative Linguistic Features. Arab World English Journal for Translation and Literary Studies. 2026;10(4):48-60. doi:10.24093/awejtls/vol10no4.4
BibTeX
@article{Peng2026,
author = {Peng, Xingwen and Duan, Huiqiong},
title = {A Comparison of Human and DeepL Translations Based on Quantitative Linguistic Features},
journal = {Arab World English Journal for Translation and Literary Studies},
year = {2026},
volume = {10},
number = {4},
pages = {48--60},
doi = {10.24093/awejtls/vol10no4.4}
}
EUVIMED