Abstract

We address the question of whether or not geographical, cognitive, and collaborative proximity have an impact on citation probability in the scientific writings of Canadian nanotechnology. Even though a number of studies in the proximity literature deal with the effects of spatial distance, scientific specialization, and social network structure, to our knowledge no one has combined all three to explore the production of academic information.

We generate a feature framework based on measurements from these factors, relying on statistical and classification approaches to assess their influence on effective citations. Specifically, by means of applying binary regression models along with tree-based machine learning algorithms, we found statistical significance proving that these features have both a verifiable impact and predictive potential. Importantly, our work is the first one that we have seen combining these techniques to infer the establishment of positive citation links. Moreover, we employed inductive network analysis comprehensively to examine the co-authorship links between authors publishing in nanoscience, considering additional network metrics to the ones usually adopted in the literature.

Our findings reveal that cognitive proximity, closely followed by the collaborative aspect, are the most important elements inducing Canadian scholars to cite, with geography sometimes acting as their base. Our results enable us to reach better understanding related to the citation behavior of the nanoresearch community in Canada, making our work a valuable contribution to scholarly literature, also giving us ground to make policy recommendations.