Discovering Structural Hole Spanners in Dynamic Networks via Graph Neural Networks

TL;DR

This paper introduces GNN-SHS, a graph neural network model designed to efficiently identify structural hole spanners in dynamic networks, outperforming existing methods in speed and accuracy.

Contribution

The paper presents the first GNN-based approach for discovering SHSs in dynamic networks, significantly improving efficiency over traditional methods.

Findings

GNN-SHS is at least 31.8 times faster than comparative methods.

GNN-SHS achieves high accuracy in identifying SHSs.

The approach effectively handles the NP-hard problem of discovering SHSs in dynamic networks.

Abstract

Structural Hole (SH) theory states that the node which acts as a connecting link among otherwise disconnected communities gets positional advantages in the network. These nodes are called Structural Hole Spanners (SHS). SHSs have many applications, including viral marketing, information dissemination, community detection, etc. Numerous solutions are proposed to discover SHSs; however, most of the solutions are only applicable to static networks. Since real-world networks are dynamic networks; consequently, in this study, we aim to discover SHSs in dynamic networks. Discovering SHSs is an NP-hard problem, due to which, instead of discovering exact k SHSs, we adopt a greedy approach to discover top-k SHSs. Motivated from the success of Graph Neural Networks (GNNs) on various graph mining problems, we design a Graph Neural Network-based model, GNN-SHS, to discover SHSs in dynamic networks, aiming to reduce the computational cost while achieving high accuracy. We analyze the efficiency of the proposed model through exhaustive experiments, and our results show that the proposed GNN-SHS model is at least 31.8 times faster and, on an average 671.6 times faster than the comparative method, providing a considerable efficiency advantage.