The Application of Cognitive Linguistics Theory in Japanese Language Teaching in the Age of Artificial Intelligence

This paper firstly summarizes the tense expressions in Japanese, Chinese and English according to the cognitive linguistics theory so as to derive the grammatical transformation law in the process of Japanese to Chinese or Japanese to English. Secondly, on the basis of artificial intelligence technology, a hierarchical Japanese phrase model based on maximum entropy tense classification features is proposed, and the extraction of Japanese phrase rules is realized through maximum entropy mathematical modeling to obtain the model feature function and decoding strategy, and the application of the model in Japanese language teaching is explained in the research. Then, the second-year students of Japanese language major in the School of Foreign Studies, Nanjing University of Posts and Telecommunications are selected as the research objects, research tools and experimental process are determined, and simulation analysis and statistical analysis are used to analyze the example of Japanese language teaching with the integration of artificial intelligence. The results show that on the model, compared with the baseline, the model in this paper improves the blue value by 0.64%, the comprehensibility by 4.39%, and the tense accuracy by 4.14% on the whole test set, which verifies the effectiveness of the hierarchical phrase model fused with the tense features proposed in this paper in Japanese-Chinese translation. In terms of Japanese language teaching, the independent samples t-test p-values were less than 0.05, and the students in class A outperformed the students in class B in the four question types: multiple choice, sentence ordering, Chinese-to-Japanese translation and composition. This study improves students’ language proficiency and achieves the objective of developing high-quality talents.