Keywords:Knowledge Transfer, Bibliometrics, Organization Management
From the aspect of the efficient operation of a research organization, it is important to manage the research platform to promote the future publishing of scientists. However, the previous ways of organization evaluations cannot distinguish the evaluation based on scientists’ community and that based on its research platform. We propose the methodology to extract scientist’s movement and research productivity from bibliographic records on Computer Science between 2007 and 2018 on Scopus and make quantitative analyses about institutional impacts on scientific innovation. As a result, scientists moving to state-of-the-art institutions does not always improve scientist’s productivity, rather collecting talented authors. Divided by its nationality, Chinese affiliations in low research productivity enhance the performance of scientists. Conversely, Japanese institutions give less contribution to scientists’ productivity. This analysis contributes to a better understanding of scientists' incentive and what research organization can do to make scientific innovation.