Expert-Driven Graph Isomorphism Networks for Key Node Identification

doi:10.1016/j.eswa.2026.134334

Expert-Driven Graph Isomorphism Networks for Key Node Identification

2026-09-09

0

OA

AI

DOI：10.1016/j.eswa.2026.134334

Original

Original request for help

Save

Abstract

En 中文

• Proposing an expert-driven GIN framework for key node identification in the SIR-model networks. • Leveraging centrality features and regression–ranking loss to enhance node importance learning. • Demonstrating superior accuracy and efficiency on 18 real-world networks in comparison with existing methods.

Keywords:

Key Node Identification

Complex Network

Graph Isomorphism Network

Susceptible-Infected-Recovered Model

Data-Driven Expert System

Expert-Driven Graph Isomorphism Networks for Key Node Identification

Abstract

Journal

Organization

Cited Papers