Developing a GPT-Based Chatbot Using Advanced Prompt Engineering for English Proficiency Level-Differentiated Reading Questions

Developing a GPT-Based Chatbot Using Advanced Prompt Engineering for English Proficiency Level-Differentiated Reading Questions

This study explores the development of an AI-facilitated chatbot for generating English reading comprehension questions, employing ChatGPT, a large language model, to meet the increasing demand for AI-enhanced English language education. Incorporating Bloom’s Taxonomy and the Common European Framework of Reference for Languages (CEFR) as foundational frameworks, the chatbot is engineered to formulate English reading questions that resonate with the varied proficiency levels of L2 learners. The study utilizes an integrative approach by combining Chain of Thought (CoT) and Few-Shot techniques in advanced prompt engineering methods to generate initial prompts for ChatGPT. Through interactions with ChatGPT, these prompts were refined, guiding the development of the chatbot. The final chatbot demonstrated its capability to produce customized English reading questions. This innovation indicates a significant advance in personalized reading instruction and assessment. It also supports the motivation for personalized learning among English language learners. This study concludes with cautious considerations and directions for future research on AI-assisted language learning tools.