In the context of rapid development of the digital economy and new quality productivity, the traditional mechanism of industry academia integration can no longer meet the precise needs of industrial upgrading and talent cultivation. Firstly, this study combines multimodal knowledge graph (MMKG) Generative big models and cross language joint learning models are used to establish a closed-loop improvement mechanism for industry academia integration, namely “demand perception resource matching field implementation evaluation and feedback iterative optimization”. Secondly, open teaching data from Chinese MOOCs, learning channels, and enterprise needs are integrated to construct a professional multimodal knowledge graph for education industry integration, and vertical domain big models are used for intelligent demand analysis and accurate resource matching. Finally, machine learning, statistical quantification, comparative experiments, and other methods are used to verify the effectiveness of the proposed industry academia integration closed-loop improvement mechanism. The experimental results show that the proposed model mechanism improves the industry academia collaboration index from 0.51 to 0.88, and shortens the demand response time by 88.2%, providing support for digital and intelligent transformation. Provide practical technical approaches and theoretical support.
