Artificial intelligence-based machine translation and human translation: English-Spanish translation of multi- word terms in scientific texts
Date Issued
2023
Author(s)
Faúndez Funes, Nicolás
Tapia Reyes, Carlos
Salazar Vergara, Brandon
García Plaza, Paulina
Abstract
The language services industry is undergoing a revolutionary merger with artificial intelligence (AI), especially in translation and scientific translation. AI is revolutionizing the accomplishment of unprecedented efficient and precise tasks. Despite challenges such as job displacement, professionals, freelancers, and students must strategically embrace AI as a transformative asset. In this study, we analyze professional human translation versus neural machine translation, specifically focusing on Chat-GPT and DeepL. We employ a multifaceted approach to delve into the intricate process of translating technical terminology from English to Spanish, specifically concentrating on the patterns "adjective + noun + noun" (situation 9) and "noun + noun" (situation 8) found in the text Cellular and Molecular Immunology. The evaluation compares the translation accuracy of ChatGPT and DeepL, highlighting the superiority of ChatGPT in chain compounds with the "n + n" pattern of multiple terms but facing challenges with "adj + n + n" terms due to the complexity of English adjectives. Despite being a recent tool, ChatGPT demonstrated commendable terminological accuracy, especially in scientific contexts, approaching human translator accuracy more closely than DeepL. In post-editing, ChatGPT proved useful for simpler structures but encountered some limitations in translating chain compounds in cellular and molecular immunology, emphasizing the need for careful examination in nuanced linguistic contexts.
File(s)![Thumbnail Image]()
![Thumbnail Image]()
Loading...
Name
bib 30.968.pdf
Size
1.33 MB
Format
Adobe PDF
Checksum
(MD5):d171b0af55accea1600665403a175623
Loading...
Name
A.30.968.pdf
Size
415.37 KB
Format
Adobe PDF
Checksum
(MD5):f5571f98b78aeddb28746530a00e0f02


