Abstract

Collaboration is a major factor in the knowledge and innovation creation in emerging science-driven industries, where the technology is rapidly changing and constantly evolving, such as nanotechnology. The scientific collaborations among individuals and organizations form knowledge co-creation network within which information is shared, innovative ideas are exchanged and new knowledge is generated. Although various simulation attempts have been carried out recently to analyze the performance of such networks at the firm level, the individual level has not been much explored in the literature yet.
The objective of this thesis is to investigate the role of individual scientists and their collaborations in enhancing the knowledge flows, and consequently the scientific production within the Canadian nanotechnology scientists. The methodology involves two main phases. First, in order to understand the collaborative behavior of scientists in the real world, the data on all the nanotechnology journal publications in Canada was extracted from the SCOPUS database and the scientists' research performance and partnership history was analyzed using social network analysis. Moreover, the predominant properties that make a scientist sufficiently attractive to be selected as a research partner were determined using data mining and through a questionnaire sent directly to the researchers selected from our database. In the second phase, an agent-based model using Netlogo has been developed to simulate the knowledge-based network where several factors regarding the ratio, existence and absence of various categories of scientists could be controlled.
It was found that scientists in centralized positions in such network have a considerable positive impact on the knowledge flows, while loyalty and cliquishness negatively affected the knowledge transmission. Star scientists appear to play a substitutive role in the network as most famous and trustable partners to be selected when usual collaborators are scarce or missing. Besides, the changes in the performance of some categories in case of the absence of others have been also observed.
The major contribution of this work stems from the fact that the developed simulation model is the first one, which is fully based on the real data and on the observed behavior of the scientists in knowledge-based network.