NEDindex: A new metric for community structure in networks

TL;DR

This paper introduces NEDindex, a new metric for evaluating community structure in networks that demonstrates consistency and effectiveness in identifying strongly connected communities across various types of networks.

Contribution

The paper presents NEDindex, a novel metric that improves upon existing measures by maintaining consistency in special cases and being computationally simple.

Findings

NEDindex shows high consistency with other metrics.

It effectively identifies communities in simulated and real networks.

Comparative analysis demonstrates its advantages over existing metrics.

Abstract

There are several metrics (Modularity, Mutual Information, Conductance, etc.) to evaluate the strength of graph clustering in large graphs. These metrics have great significance to measure the effectiveness and they are often used to find the strongly connected clusters with respect to the whole graph. In this paper, we propose a new metric to evaluate the strength of graph clustering and also study its applications. We show that our proposed metric has great consistency which is similar to other metrics and easy to calculate. Our proposed metric also shows consistency where other metrics fail in some special cases. We demonstrate that our metric has reasonable strength while extracting strongly connected communities in both simulated (in silico) data and real data networks. We also show some comparative results of our proposed metric with other popular metric(s) for Online Social Networks (OSN) and Gene Regulatory Networks (GRN).