A Study of Human-AI Interaction Patterns in Artificial Intelligence-Assisted Second Language Writing

The aim of this project is to investigate human-computer interaction patterns in AI-assisted second language writing in order to explore students' strategies and patterns when collaborating with generative AI tools, as well as key hidden states in AI-assisted writing. The project utilizes Hidden Markov Models (HMM) and process mining techniques to analyze the potential patterns of student interactions with AI and evaluate the efficacy of different patterns. The findings will reveal the cognitive mechanisms of human-computer interaction in second language writing, expand the theoretical framework of human-computer collaboration, and provide empirical evidence for effectively guiding students to utilize AI to address key questions in second language writing research.