Constructing and Validating an AI-Empowered Evaluation Framework for Industry–Education Integration

In response to the problems of scattered indicators, insufficient result verification, and weak interpretability in the evaluation of industry education integration, this article constructs an artificial intelligence (AI) empowerment evaluation framework. Based on observations from 126 universities and 3010 questionnaires from 42 universities from 2022 to 2024, this study integrates five dimensions: collaborative governance, curriculum co construction, practical platform, process quality, and outcome effectiveness. Research combines Analytic Hierarchy Process (AHP), Entropy Weight Method, and Light Gradient Boosting Machine (LightGBM) modeling. The results showed that the mean absolute error (MAE) and root mean square error (RMSE) of the LightGBM model were 0.138 and 0.189, respectively, with a coefficient of determination (R ²) of 0.814; The framework score has a significant positive impact on the corresponding employment rate, enterprise satisfaction, and collaborative innovation output. The key driving factors for high-quality industry education integration are the participation intensity of enterprise mentors, the coverage rate of curriculum co construction, and the proportion of dual teacher teachers, indicating that the deep participation of enterprises in curriculum and practice is more important.