The vulnerability of communities in complex network: An entropy approach

TL;DR

This paper introduces an entropy-based method to measure community vulnerability in complex networks, integrating internal and external factors for more accurate and comprehensive assessments.

Contribution

It proposes a novel entropy approach that combines multiple community factors, addressing limitations of existing vulnerability measures in complex networks.

Findings

The method effectively evaluates community vulnerability in real-world networks.

Experimental results show the proposed approach outperforms existing methods.

Sensitivity analysis confirms robustness of the measure.

Abstract

Measuring the vulnerability of communities in complex network has become an important topic in the research of complex system. Numerous existing vulnerability measures have been proposed to solve such problems, however, most of these methods have their own shortcomings and limitations. Therefore, a new entropy-based approach is proposed in this paper to address such problems. This measure combines the internal factors and external factors for each communities which can give the quantitative description of vulnerability of community. The internal factors contain the complexity degree of community and the number of edges inside the community, and the external factors contain the similarity degree between chosen community and other communities and the number of nodes outside the community. Considering community vulnerability from the perspective of entropy provides a new solution to such problem. Due to sufficient consideration of community information, more reasonable vulnerability result can be obtained. In order to show the performance and effectiveness of this proposed method, one example network and three real-world complex network is used to compare with some exiting methods, and the sensitivity of weight factors is analysed by Sobol' indices. The experiment results demonstrate the reasonableness and superiority of this proposed method.