Abstract
The purpose of this study is to empirically demonstrate the effects of feedback based on AI automatic assessment. The findings of this study demonstrate that the introduction of an AI automatic assessment system significantly contributes to improving L2 learners' listening comprehension abilities from three perspectives: (1) providing real-time, individualized, and continuous feedback; (2) detailed visualization of the listening comprehension process and elucidation of cognitive processes; and (3) qualitative understanding of comprehension content through importance-based assessment and formulation of appropriate instructional plans. A particularly important finding is that recall rates by importance level enable multilayered assessment of the quality of learners' listening comprehension.