On a Triadic Approach to Connect Microstructural Properties to Social Macrostructural Patterns

Abstract—Although extensive studies have been conducted on revealing the global statistical characteristics of online social networks (OSNs), such as degree distribution and diameters, little attention has been paid to investigating social macrostruc-tural properties of the network. Social macrostructures, such as structural balance, ranked clusters and transitivity, are of great importance on account of their abilities to reflect the underlying social psychological processes related to the formation and evolution of relationships among people in our society. Here we present a detailed study on examining the existence and evolution of social macrostructural patterns in an em-pirical online social network, and exploring how they can be explained and predicated by network microstructural properties, i.e. nodal outdegree and indegree distribution and dyad census by analyzing the network triadic structures. We introduce a novel clustering coefficient based network sampling approach to efficiently examine triad distribution in online graphs. It is shown that the distribution of observed triad census is low dimensional and can be greatly explained by network dyadic properties in our data. In a time series analysis, we observe that the network under study exhibits strong tendencies towards balanced, transitive and clustered social structure given the network nodal and dyadic characteristics over time. Our findings supplement the studies on structural properties of online social network by providing more insights on the relation between network macrostructures and the micro-level social processes that result in them, and form the basis to understand better how online social media systems change the information and communication fabric of our society. I