DynamicScore: a Novel Metric for Quantifying Graph Dynamics

TL;DR

This paper presents DynamicScore, a new metric for quantifying the evolution of graphs by considering both the changes in vertices and edges and their composition, applicable to various types of dynamic networks.

Contribution

The paper introduces DynamicScore, a novel metric that captures detailed graph dynamics, including set composition, for evolving networks like preferential attachment and Edge-Markovian graphs.

Findings

DynamicScore effectively measures graph changes over time.

Application to different models reveals insights into network evolution.

Results enhance understanding of dynamic graph behavior.

Abstract

This study introduces a new metric called ''DynamicScore'' to evaluate the dynamics of graphs. It can be applied to both vertices and edges. Unlike traditional metrics, DynamicScore not only measures changes in the number of vertices or edges between consecutive time steps, but also takes into account the composition of these sets. To illustrate the possible contributions of this metric, we calculate it for increasing networks of preferential attachment (Barab{\'a}si-Albert model) and Edge-Markovian graphs. The results improve our understanding of the dynamics inherent in these generated evolving graphs.