Modeling knowledge diffusion in scientific innovation networks: an institutional comparison between China and US with illustration for nanotechnology

Knowledge is a crucial asset in organizations and its diffusion and recombination processes can be affected by numerous factors. This study examines the influence of the status of individual researchers in social networks on the knowledge diffusion and recombination process. We contend that knowledge diversity, random diffusion, and parallel duplication are three primary factors characterizing diffusion paths in knowledge networks. Using multiple social network measures, we investigate how individuals in the respective institutional collaboration networks influence knowledge diffusion through scientific papers. Scientific publication data and citation data from six prolific institutions in China (Chinese Academy of Sciences) and the United States (University of California at Berkeley, University of Illinois, Massachusetts Institute of Technology, Northwestern University, and Georgia Institute of Technology) in nanotechnology field in the interval 2000–2010 were used for empirical analysis, and the Cox regression model was leveraged to analyze the temporal relationships between knowledge diffusion and social network measures of researchers in these leading institutions. Results show that structural holes and degree centrality are the most effective measures to explain the knowledge diffusion process within these six institutions. Knowledge recombination is mainly achieved through parallel duplication within groups and recombination of diverse knowledge across different groups. The results are similar for all six institutions except for Bonacich power and eigenvector measures, which may posit cultural difference across countries and institutions.