Understanding Geographical Patterns of Scientific Collaboration in Artificial Intelligence among Canadian Researchers

The role of geographical proximity in facilitating international scientific collaboration has been studied thoroughly in recent years. However, the effect of geography and its interaction with network proximity on forming scientific collaborations among researchers who work under the same national systems has not been appropriately addressed. Using co-publication data of Canadian AI scientists from 2000 to 2019, we studied the relationship between different forms of geographical proximity and national scientific collaboration in Canada. The logit regression results show that geography is still an important impediment to national scientific collaboration despite the tremendous improvements in transportation and communication technologies during recent decades. Besides, we found a substitutability pattern between network proximity and some forms of geographical proximity in national collaborations.