Data compression to choose a proper dynamic network representation

TL;DR

This paper introduces a data compression-based method to select the most suitable dynamic network representation among snapshots, link streams, and interval graphs, enhancing analysis and storage efficiency.

Contribution

It proposes a novel approach using data compression to objectively choose the best dynamic network representation for a given dataset.

Findings

The method effectively distinguishes the most appropriate representation.

Application on datasets demonstrates improved storage and analysis efficiency.

The approach is applicable to both synthetic and real-world data.

Abstract

Dynamic network data are now available in a wide range of contexts and domains. Several representation formalisms exist to represent dynamic networks, but there is no well-known method to choose one representation over another for a given dataset. In this article, we propose a method based on data compression to choose between three of the most important representations: snapshots, link streams and interval graphs. We apply the method on synthetic and real datasets to show the relevance of the method and its possible applications, such as choosing an appropriate representation when confronted to a new dataset, and storing dynamic networks in an efficient manner.