Journal of Studies in Language, Culture and Society (JSLCS)
Volume 8, Numéro 2, Pages 271-293
2025-06-30
From Prompt To Response: Understanding Chatgpt Language Processing Framework To Improve Efl Teaching
Authors : Bengrait Nawel .
Abstract
Foreign Language (FL) learning and teaching have known a drastic shift due to the implementation of the fast-growing Artificial Intelligence (AI) tools. Traditional methods to teaching English involve one-size-fits-all approach that makes it challenging to address individual needs. Teachers interacting with AI-language assistant chabots as ChatGPT for generating teaching materials and designing tasks that align with students’ different proficiency levels, may not receive the intended response of the query. This can be the result of their limited knowledge of how the model function to generate texts. The present study analyzes ChatGPT language processing framework, centered on key components of prompting, tokenization, vectorization, transformer mechanism, and semantic analysis. The research aims to analyze how ChatGPT interprets and generates human-like texts, and hypothesizing that these stages can impact significantly the response coherence and accuracy. Adopting a qualitative method, the study illustrates language processing features through examples, showcasing its capacity to extract meaning from written text, and the importance of context in constructing responses. The Findings revealed that for the Large Language Model (LLM) to provide the accurate response to the given prompt, it interacts with the input lexical and semantic structures, and contextual specificity. EFL teachers understanding prompt engineering can refine their queries to receive more reliable responses which permit them to master the skill to lessen ambiguity in prompts generation and improve AI interaction for academic and educational purposes.
Keywords
Artificial Intelligence ; ChatGPT ; context ; English as a Foreign Language ; Large Language Model ; prompt ; response
Les articles similaires
Dou Abdelbasset
.
Rim Kelthoume
.
pages 64-75.
Bentaleb Ibtissam
.
pages 44-56.
