Kempten Autumn Talks

Title of the lecture

“Expert selection in networks”

Abstract

The Top Candidate algorithm was developed to identify experts on recommendation networks. It relies on a simple observation: experts are much more effective in identifying other experts. In this talk we show two applications of this method. We rank academic institutions by their performance in a scientific discipline. We compile a citation network from game theoretic literature published between 2008-2017. The Top Candidate algorithm not only shows the current position of an institution but also reveals how difficult it is to improve it. In our second case study we look at innovation spreading in social networks. Early adopters play an important role in the innovation diffusion process. We show that the Top Candidate method identifies innovators and early adopters much more efficiently than other centrality measures.