Using Agent-Based Modeling and Dynamic Data Envelopment Analysis to Measure the Knowledge Diffusion Efficiency in the Project Portfolio Network

Aiming at the problem of knowledge diffusion between projects, the paper uses the agent-based modeling and multi-stage DEA to analyze the dynamic process of knowledge diffusion, so as to more accurately measure the impact of knowledge diffusion on project portfolio performance. Firstly, taking each project in the portfolio network as an agent, the relationship of knowledge diffusion and feedback between projects as the connection of network, the paper constructs the dynamic model of knowledge diffusion between projects, and analyzes the impact of knowledge diffusion on the knowledge transmission probability and knowledge level. Further, based on the analysis of knowledge diffusion between projects, the measurement model of knowledge transmission probability and knowledge level considering knowledge feedback and propagation is constructed. Finally, based on the multi-stage DEA model, the impact of knowledge diffusion on portfolio performance is measured. Finally, a research and development project portfolio are taken as an example in order to demonstrate the effectiveness of the method presented in this research