Abstract:In the context of an innovation ecosystem, node enterprises utilize deep network embedding advantages to share knowledge with participating entities, which becomes the key to improving their knowledge innovation performance, establishing innovation competitive advantages, and enhancing the knowledge ecological performance of the innovation ecosystem. Based on the resource-based view and social network theory, construct a theoretical model of the impact mechanism of knowledge innovation performance in node enterprises; Introduce specific indicators to characterize the process of knowledge sharing, innovation knowledge investment, network embedding, and construct a system dynamics model for knowledge innovation performance of node enterprises. Explore the dynamic impact mechanism of knowledge sharing within the innovation ecosystem on the knowledge innovation performance of node enterprises and the dynamic regulatory role of network embedding in the impact mechanism. Research has found that high-level network embedding is beneficial for improving the knowledge innovation performance of node enterprises and the knowledge ecosystem performance of innovation ecosystems; The slowing down of the growth rate of innovative knowledge leads to a marginal diminishing effect on knowledge innovation performance; The improvement of the level of institutional embedding is conducive to weakening the marginal diminishing effect of knowledge innovation performance; Reasonably increasing knowledge innovation investment is beneficial for improving knowledge innovation performance. Based on this, policy recommendations are proposed to optimize the knowledge sharing process of the innovation ecosystem, dynamically optimize the level of institutional embedding in the innovation ecosystem of node enterprises, dynamically adjust the proportion of knowledge innovation investment, and flexibly match knowledge partners.