The integration of interdisciplinary knowledge in higher education teaching has gradually become an important way to improve teaching quality and innovation ability. This paper proposes a method of designing and realizing the interdisciplinary knowledge integration and teaching innovation system of dual colleges and universities based on artificial intelligence algorithms. The study constructs a multidimensional disciplinary knowledge network through graph convolutional self-encoding model and LDA topic model, and analyzes the knowledge fusion process between dual colleges and universities. The experimental results show that among the 15 themes, the contribution degree of theme 2 “Big Data and Artificial Intelligence Application” is 0.9878, which indicates its important position in the interdisciplinary knowledge integration between dual universities. The point centrality analysis of the knowledge network shows that “interdisciplinary” and “artificial intelligence” are the most influential knowledge nodes. This method not only provides theoretical support for interdisciplinary knowledge integration in dual colleges and universities, but also provides practical basis for teaching innovation in related fields. This paper shows that the integration of interdisciplinary knowledge through artificial intelligence algorithms can effectively promote the quality of teaching and research innovation in universities, and enhance the overall competitiveness of universities.
