Effective graph-neural-network based models for discovering Structural Hole Spanners in large-scale and diverse networks

Download PDF
Open Webpage

Diksha Goel, Hong Shen, Hui Tian, Mingyu Guo

Published: 01 Jan 2024, Last Modified: 17 Dec 2024Expert Syst. Appl. 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Aims to discover SHS nodes in large-scale and diverse networks.•Introduces GraphSHS, a graph neural network model to discover SHSs in large networks.•Proposes Meta-GraphSHS to discover SHSs across diverse networks.•GraphSHS demonstrates a minimum 167.1x speedup over baselines in real-world networks.•Meta-GraphSHS achieves 96.2% accuracy in identifying SHSs across diverse networks.