Research on the Construction and Application of New Business Education Integration Teaching Model Empowered by Artificial Intelligence

Abstract

New business education requires a teaching model capable of dynamically aligning curriculum content, industrial skill demand, project allocation, and practical decision training. To reduce the mismatch between fast-changing industrial requirements and static curriculum revision cycles, this study constructs an AI-empowered industry–education integration model. A dynamic industry-skill knowledge graph is built from 45,600 job postings using BERT and BiLSTM-CRF to extract job-skill entities and relationships. A semantic mapping mechanism calculates the alignment between course syllabi and industry skill nodes, producing curriculum update suggestions. A multi-agent reinforcement learning scheduler then matches students and enterprise projects using ability vectors, project demand vectors, and reward functions considering semantic fit, team complementarity, and resource idleness. Finally, a generative business simulation sandbox based on cGAN and large language models creates dynamic market demand and competitor strategies for decision training. A 16-week controlled experiment shows that AI-driven mapping raises semantic alignment to 0.89, reinforcement learning reduces project-resource idle rate to 5.7%, and generative sandbox training improves students’ strategy innovation and risk-prediction scores by more than 23 points. The model provides an intelligent decision-support framework for adaptive education systems and industry–education resource coordination.