AI-enabled Multimodal Feedback for Chinese as a Second Language Learners

Abstract

Recent research has shown that AI-enabled multimodal feedback is effective and meaningful for language learners. However, there is limited understanding of how to support students, particularly young learners, in interacting with such feedback to reflect on and improve their learning. This poster presents the design of ARChE 2.0, an AI-enhanced Chinese language learning system tailored for lower primary students in Singapore. The system provides AI-generated multimodal feedback, including automatic star ratings, audio-based suggestions on grammar and content, and visual feedback in the form of AI-generated images, to support both vocabulary learning within and beyond the classroom. Preliminary results from the ongoing intervention suggest that AI-enabled multimodal feedback can effectively promote continuous improvement in students' learning outcomes during collaborative tasks. This study contributes to the growing field of AI in education by offering insights into the design of engaging, feedback-rich learning environments for young learners in language learning. Future research will further explore how lower primary students actively interpret, respond to, and act upon multimodal feedback, and identify pedagogical strategies that support students in self-regulating and processing the feedback more effectively.