Abstract: Generative artificial intelligence has been deeply integrated into Japanese language teaching in universities, reshaping its teaching paradigm, but the problem of cultural adaptability is prominent. Large language models represented by DeepSeek have problems of cultural distortion and emptiness in Japanese honorifics, high-context expression, cultural value presentation and non-verbal communication elements, which are likely to mislead learners and hinder the cultivation of intercultural communication competence. This paper constructs a dual-core framework of "Cultural Filter + Scenario Enhancement", calibrates AI output and upgrades teaching through knowledge graphs, model fine-tuning, etc., clarifies that teachers should transform into cultural arbitrators, and provides theoretical and practical references for human-machine collaborative Japanese teaching and the cultivation of talents with both language ability and cultural insight.
齐金玲. 人工智能辅助高校日语教学中的文化适配性问题与解决路径研究[J]. 吉林化工学院学报, 2026, 43(2): 23-27.
QI Jinling. Research on Cultural Adaptation Issues and Solutions in AI-Assisted Japanese Language Teaching in Higher Education. Journal of Jilin Institute of Chemical Technology, 2026, 43(2): 23-27.
