Scientific Development Research

In the digital classroom environment, the improvement of fluency in spoken Chinese as a second language faces two core issues: limited teaching interaction and insufficient technological application. At the teaching level, the interactive mode is single, lacking real situations and feedback; On a technical level, the accuracy of speech recognition, the degree of multimodal fusion, and the incomplete application of large-scale language models are not high. In response to the above issues, this article proposes to use a large language model as the core driving force to optimize the path from four aspects: construction of multimodal immersive training scenarios, optimization of personalized real-time feedback systems, construction of hierarchical advanced training frameworks, and monitoring and optimization mechanisms for learning data. The aim is to systematically improve the level of Chinese second language teaching and provide an integrated solution for enhancing learners' oral fluency.

Big language model; Multimodal interaction; Digital classroom; Chinese second language learners; Oral fluency