Exploring the role of artificial intelligence in facilitating writing assessment in second language learning

• In Event: SICET - AI & Assessment

Short Description

This study explored the role of artificial intelligence (AI) in assessing writing accuracy, particularly at the T-unit and sentence levels. Large language models (LLMs), such as ChatGPT, iFLYTEK, and Baidu Cloud, were evaluated for their robustness and efficiency in identifying errors in writing. The study used a confusion matrix and evaluation metrics to compare the LLM results with human rating results. Qualitative analysis was performed to analyze the discrepancies between human and AI assessment results.

Authors

• Contributor: Zilu Jiang, The Ohio State University

• Contributor: Zexin Xu, The Ohio State University

• Contributor: Zilong Pan, Lehigh University

• Contributor: Jingwen He, The Ohio State University

• Contributor: Kui Xie, Michigan State University