Design and Development of an AI-Powered IELTS Writing Assessment System for Task 2 Essays

This study aims to design and develop an AI-based scoring system for IELTS Task 2 writing and ensure the validity and practicality of its scoring criteria through expert consensus. The research employed the Delphi Method, involving 10 experts from the fields of language teaching and educational technology. Based on the requirements of the scoring system design and expert feedback, a critical indicator framework for the AI-based IELTS Task 2 writing scoring system was established. The consensus rates (Cm) for both rounds were 100%, and the average reliability coefficient (Cr) was 0.8. The Kendall's W coefficients for the two rounds were 0.626 and 0.712 (P < 0.001). Through two rounds of expert surveys and scoring reviews, a key design indicator framework was developed, including 7 modules and 21 functional points. This study establishes the AI-based IELTS writing scoring system, offering significant theoretical support and practical insights for the application of AI technology in language education.