Neural machine translation (MT) replicates human brain networks (Goto, 2017) and is often considered a “black box” due to its unexplained structure and function. To use MT effectively, target language (TL) learners should pre-edit source language (SL) documents, similar to how humans paraphrase L1 texts before translating to L2 (Tsuji, 2024).
Previous studies have shown that pre-editing SL texts improves the quality of MT output (Feifei et al., 2022; Kokanova et al., 2022). However, the specific differences between with and without editing MT translations have been less explored.
This study aims to compare MT translations with and without pre-editing and examine the impact of pre-editing on the quality of target-language texts. The study involved 131 Japanese university students with intermediate English proficiency, each composing a Japanese essay which was translated into English using MT. Three language researchers compared and analyzed the MT translations. They systematically classified detected issues, primarily lexical and grammatical.
The results showed that MT translations without pre-editing displayed errors that were not present in MT translations with pre-editing. The most common errors specific to MT translations without pre-editing were ‘inappropriate subject use’ and ‘unclear meaning’. Pre-edited MT translations were generally comprehensible and contained minor errors (‘inappropriate use of language’) that did not significantly affect comprehensibility.
This study indicated that pre-editing source texts is essential for improving the quality of MT translations. The tips for pre-editing could potentially be used as strategies of L1 paraphrasing required for the human L2 translation process.

Authors:
Kayo Tsuji, Osaka Metropolitan University, Japan
Kiyo Okamoto, Osaka Metropolitan University, Japan
Benjamin Neil Smith, Osaka Metropolitan University, Japan

About the Presenter(s)
Kiyo Okamoto is a lecturer at Osaka Metropolitan University in Japan.

See this presentation on the full schedule – Friday Schedule

Bookmark

---

A Note to Presenters

To enhance academic profiles and showcase research, we encourage all presenters and co-presenters to include links to their public LinkedIn, ResearchGate profile, and research websites. Presenters may update their bio for their presentation by completing the form linked below by October 22, 2024.
- Presenter Information Update Form
Submitted changes will be reflected on November 01, 2024

Additionally, presenters should also update their IAFOR account details if there have been any changes to affiliations or biographies.
- https://submit.iafor.org/my-account/edit-account