Implementation of a Learning Support System Based on ChatGPT for Improving English Reading Comprehension Abilit

Abstract

This study presents the design and implementation of a learning support system that leverages ChatGPT to improve reading comprehension of English passages in secondary education. Traditional English reading practice often requires significant effort for material preparation and dictionary use, which can hinder learning efficiency. To address this issue, the proposed system, developed on Pythonista3, automatically generates English passages and related questions by grade level and provides interactive Q&A functions to reduce the burden of dictionary searches, thereby enhancing learning efficiency. An evaluation experiment was conducted with five junior high school students. The results indicated that the system was effective in terms of usability and improving learning efficiency. However, issues were identified, including inconsistencies in text difficulty and the frequent occurrence of unfamiliar vocabulary. These findings demonstrate both the potential and limitations of applying generative AI in educational contexts. Future work will focus on refining prompt design and implementing difficulty control mechanisms, aiming to enhance the system’s practicality in classrooms and contribute to the sustained improvement of learners’ English proficiency.