Sziklai Balázs
(Magyar Tudományos Akadémia/Budapesti Corvinus Egyetem)
„How to identify experts in a community?”
Befogadó folyóirat: International Journal of Game Theory
A megjelenés ideje: 2017. június 14.
The group identification literature mostly revolves around the problem ofidentifying individuals in the community who belong to ethnic or religious groups.Here we use the same model framework to identify individuals who play key rolein some sense. In particular we will focus on expert selection in social networks.Ethnic groups and expert groups need completely different approaches and differenttype of selection rules are successful for one and for the other. We argue that stabil-ity is a key property in expert selection. The idea is that experts are more effectivein identifying each other, thus the selected individuals should support each other’smembership. We propose a parametric algorithm based on the so called top candidaterelation. The parameter expresses how permissive we want to be in expert selection.The two limit cases are the stable set and the top candidate core. The former containsvirtually everybody that can be considered as an expert, while the latter consists ofthe elite. We establish an axiomatization to show that the algorithm is theoreticallywell-founded. Furthermore we present a case study using citation data to demonstrateits effectiveness. We compare its performance with classical centrality measures.