How to facilitate knowledge diffusion in collaborative innovation projects by adjusting network density and project roles

The mechanism of knowledge diffusion in collaborative innovation projects has long been a controversial topic, partly due to the lack of attention to actors. The characteristics of actors in projects directly affect individual diffusion behaviour and, in turn, the whole process of knowledge diffusion. The interactions among actors in knowledge diffusion can be abstractly expressed as a network. In prior research, the network characteristics related to knowledge diffusion are confined to the dimensions of network relationships, and the attributes of network nodes have received limited attention. For this reason, this paper focuses on two characteristics, network density and project roles, which are argued to have a considerable correlation with knowledge diffusion in collaborative innovation projects. As knowledge diffusion in a project is a complex and dynamic process, an agent-based modelling approach was employed to construct a simulation model of knowledge diffusion. A case study was conducted that included three parallel simulation experiments with different network characteristics. The results show that (a) the adjustment of network density within a specific range is positively correlated with the extent of knowledge diffusion; (b) the role division of network nodes has a negative impact on the overall extent of knowledge diffusion; and (c) the role division of network nodes has a particular moderating effect on the relationship between density and the extent of diffusion. This research reveals the mechanism of knowledge diffusion in collaborative innovation projects, which provides theoretical guidance for designing the relational structure and the roles of actors in practice.