Likelihood-based approach to discriminate mixtures of network models that vary in time

TL;DR

This paper introduces a likelihood-based method to identify when and how the underlying mechanisms of network growth change over time, using both artificial and real network data.

Contribution

It presents a novel likelihood framework for detecting change points in network growth models and analyzing the evolution of underlying mechanisms.

Findings

Successfully recovered change points in artificial networks.

Demonstrated changing importance of growth mechanisms in real datasets.

Framework can identify temporal shifts in network formation processes.

Abstract

Discriminating between competing explanatory models as to which is more likely responsible for the growth of a network is a problem of fundamental importance for network science. The rules governing this growth are attributed to mechanisms such as preferential attachment and triangle closure, with a wealth of explanatory models based on these. These models are deliberately simple, commonly with the network growing according to a constant mechanism for its lifetime, to allow for analytical results. We use a likelihood-based framework on artificial data where the network model changes at a known point in time and demonstrate that we can recover the change point from analysis of the network. We then use real datasets and demonstrate how our framework can show the changing importance of network growth mechanisms over time.