A graph neural network-based role classification in criminal networks CSCI–RTCW

Dogan, Vedat; Prestwich, Steven D.; O'Sullivan, Barry

A graph neural network-based role classification in criminal networks CSCI–RTCW

Files

A Graph Neural Network based Role Classification in Criminal Networks.pdf(576.96 KB)

Accepted Version

Date

2024-12-12

Authors

Dogan, Vedat

Prestwich, Steven D.

O'Sullivan, Barry

Abstract

Understanding the roles of individuals in terrorist networks is an important task in counter-terrorism. This paper presents the first application of graph neural networks to this task. We apply our approach to a real-world terrorist network representing three different ideologies and nine specific groups. We demonstrate the challenges associated with this task and present the framework using graph neural networks and their advantages in this context.

Keywords

Terrorist networks , Counter-terrorism , Role identification , Graph neural networks , Node classification

Citation

Dogan, V., Prestwich, S. and O'Sullivan, B. (2024) 'A graph neural network-based role classification in criminal networks CSCI–RTCW', 2024 International Conference on Computational Science and Computational Intelligence (CSCI), Las Vegas, NV, USA, December 11-13, 2024.