On a pyramid structure in social networks

Wenruo Lyu,Liang Zhao,Zhao, Liang

DOI: https://doi.org/10.1007/s13278-024-01286-x

2024-07-03

Social Network Analysis and Mining

Abstract:This study introduces a hierarchical pyramid structure as a novel framework for social network analysis, differing fundamentally from traditional hierarchical models by its expansive, connected topology from the apex downwards. We undertake a comprehensive examination of key social network attributes, such as degree distribution, clustering coefficients, and small-world properties, through this new lens. Our empirical investigation supports a hypothesis regarding the predominance of large pyramids in real-world social networks, corroborating findings from the existing literature. Notably, our results reveal significant variations in the size distribution of basic pyramids across different types of networks. This disparity indicates that current generative models for social networks may not adequately capture the complexities inherent in real-world interactions, thus underscoring the need for innovative modeling techniques. Moreover, this paper proposes a pyramid-based criterion for defining the small-world characteristic within individual networks , a departure from conventional metrics that typically generalize across network types . By applying this framework, we conduct detailed analyses of a broad spectrum of real-world networks, validating the efficacy and adaptability of the pyramid structure in advancing social network research.