Temporal Network Pattern Identification by Community Modelling

Xubo Gao,Qiusheng Zheng,Didier A Vega-Oliveros,Leandro Anghinoni,Liang Zhao

doi:10.1038/s41598-019-57123-1

Abstract

Temporal network mining tasks are usually hard problems. This is because we need to face not only a large amount of data but also its non-stationary nature. In this paper, we propose a method for temporal network pattern representation and pattern change detection following the reductionist approach. The main idea is to model each stable (durable) state of a given temporal network as a community in a sampled static network and the temporal state change is represented by the transition from one community to another. For this purpose, a reduced static single-layer network, called a target network, is constructed by sampling and rearranging the original temporal network. Our approach provides a general way not only for temporal networks but also for data stream mining in topological space. Simulation results on artificial and real temporal networks show that the proposed method can group different temporal states into different communities with a very reduced amount of sampled nodes.

Highlights

In real-world applications, data often arrives in streams and must be analyzed in real-time
We focus on the problem of concept drift in large data-sets, which are represented by temporal networks or data that can be transformed as well
A similar problem found in the complex network literature is the change point detection[22,23,24,25], which seeks to detect the moments of change between one concept to the other

Summary

Introduction

In real-world applications, data often arrives in streams and must be analyzed in real-time. Large systems usually are high-dimensional and heterogeneous in nature, which require a discriminatory, while interactive, structural representation of the organization In this way, to better improve our understanding of such complex systems, complex network theory has become essential to move beyond simple graphs and to take such multi-type interaction features into account, which include multiple subsystems and layers of connectivity. A similar problem found in the complex network literature is the change point detection[22,23,24,25], which seeks to detect the moments of change between one concept (or network pattern) to the other This approach focus on measuring the dissimilarity between one snapshot of the temporal network and the one in time. The authors did not consider the case of multiple edges placed at any given time, and they focused on inferring a model that reproduces the waiting time in the empirical data

Methods

Results

Conclusion