Teaching Arabic-Korean translation using ChatGPT

Abstract

This study aims to examine a three-stage instructional design for an Arabic-Korean translation class, using ChatGPT as a comparison model. The study seeks to provide an effective instructional design for a translation class that adapts to the rapid development of AI-based machine translation. In this design, students are guided and supervised to practice translation and discuss strategies without relying on AI-based tools in the classroom. ChatGPT's output is used as a comparison model for their own translations. The findings indicate that this approach enhances students' critical thinking regarding AI-based machine translation and raises their awareness of the potential risks associated with uncritical reliance on such technologies without subsequent revision. However, analysis and evaluation proved to be challenging skills, as students are not accustomed to them. Further research is needed to refine this instructional design, taking into account the difficulty levels of the source texts. Additionally, tasks such as ''evaluate'' and ''create an improved version'' should be simplified and clarified, by providing more detailed steps and examples in task instructions to help students better understand the requirements.

Keywords

• Machine translation
• ChatGPT
• Korean
• Arabic