IELTS Writing Tasks Evaluation Using LLMs: A Comparative Study of ChatGPT, Claude, Copilot, and Google Gemini
Year: 2024
Type: conference-paper
Abstract: IELTS writing tasks are considered one of the most challenging tests among the other three IELTS sections: speaking, reading, and listening. The IELTS Writing section requires significant improvement effort, and obtaining expert feedback can be costly. In recent years, large language models (LLMs), a type of artificial intelligence, have revolutionized and are used in various tasks such as text summarization, question answering, and many others. This paper explores the effectiveness of wellknown LLMs in estimating IELTS Writing band scores, offering a comparative analysis of several models, including ChatGPT, Claude, Copilot, and Google Gemini. The data used for this task were collected from IELTS practice test books, and the results were evaluated using the R-square and the average degree of bias. The findings indicate that ChatGPT-4 and ChatGPT-40 showed the closest results to the actual labels and outperformed the other models.
Institution: Nahrain University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4408258390