Heterogeneous node copying from hidden network structure

Abstract

Node duplication is an established model of network formation, whereby an existing node is duplicated, and edges are formed with uniform probability to the neighbours of the duplicated node. Here, the author proposes a copying model where links are copied depending on hidden interactions between nodes, and shows analytically and numerically that this leads to higher network clustering than in the uniform copying case, a property that is also found in real collaboration networks.