Opinion Divergence Reveals the Hierarchical and Overlapping Community Structure in Networks

TL;DR

This paper introduces a novel method to detect hierarchical and overlapping communities in networks by analyzing opinion divergence during consensus processes, revealing new insights into network modular structures.

Contribution

It proposes a new approach linking opinion divergence to community structure, validated through experiments on real and artificial networks.

Findings

Opinion divergence correlates with hierarchical and overlapping community structures.

The method effectively identifies communities with high robustness and efficiency.

Connections between opinion dynamics and network modularity are established.

Abstract

Often exhibiting hierarchical and overlapping structures, communities or modular groups are fundamental and complex in network science. One of the most exploited tools to detect the mesoscopic structure is synchronization. Several phenomena including convergence rate and local convergence under constraints are studied to uncover the existence and features of communities. Here, employing a background of opinion dynamics, we study the divergence of agents' opinions, i.e. the state differences of nodes, in the progress of reaching global consensus and then reveal its connections with hierarchy and overlap in the modular structure. Furthermore, based on the obtained close relationships, a new method is proposed to identify hierarchical and overlapping communities, whose robustness and efficiency are validated via experiments on real and artificial networks. Both the connections and approaches provide a novel insight on the features of modular structures in networks.