Quantification and Comparison of Degree Distributions in Complex Networks

TL;DR

This paper introduces a new method for quantifying and comparing degree distributions in complex networks, enabling more accurate similarity assessments across networks of different scales.

Contribution

A novel quantification technique and a new distance function for degree distributions that outperform existing methods in accuracy.

Findings

Effective comparison of networks with different scales

Outperforms state-of-the-art methods in accuracy

Provides a fixed-length feature vector for degree distributions

Abstract

The degree distribution is an important characteristic of complex networks. In many applications, quantification of degree distribution in the form of a fixed-length feature vector is a necessary step. On the other hand, we often need to compare the degree distribution of two given networks and extract the amount of similarity between the two distributions. In this paper, we propose a novel method for quantification of the degree distributions in complex networks. Based on this quantification method,a new distance function is also proposed for degree distributions, which captures the differences in the overall structure of the two given distributions. The proposed method is able to effectively compare networks even with different scales, and outperforms the state of the art methods considerably, with respect to the accuracy of the distance function.