Graph Neural Networks for Supply Chain Risk Propagation
DOI:
https://doi.org/10.15662/IJRAI.2021.0405001Keywords:
Graph Neural Networks (GNNs), Supply Chain Risk Propagation, Graph Convolutional Networks (GCN), Graph Attention Networks (GAT), Supplier Dependency Networks, Network-Based Risk Modeling, Interpretability, ScalabilityAbstract
Supply chains form complex, interconnected networks where disruptions—such as supplier failures, geopolitical events, or natural disasters—can trigger cascading risks throughout the system. Traditional models often lack the capacity to capture these multi-hop dependencies and dynamic propagation effects. Graph Neural Networks (GNNs), which aggregate information across network structures, offer a powerful alternative for modeling risk propagation. This paper surveys pre-2020 work that leverages GNNs to represent supply chain entities as nodes and their interdependencies as edges, enabling prediction of systemic risks. Key methodologies include Graph Convolutional Networks (GCNs) and Graph Attention Networks (GATs), which support modeling of relationships and varied influences among supply chain actors Wikipedia. GNNs enhance understanding of how risk travels across tiers and help identify critical vulnerability points. Though early applications—such as supplier selection and procurement optimization—did not explicitly target risk propagation, they laid the foundation by evaluating relational risks and dynamic adaptation in supply networks ResearchGate. Challenges of applying GNNs in this domain include scalability, interpretability, and data availability. Addressing these issues is essential for deploying GNN-driven risk analysis in real-world supply chain management. This paper synthesizes relevant architectural approaches, discusses strengths and limitations, and outlines future directions anchored in pre-2020 foundations.
References
1. Kipf, T. N., & Welling, M. (2017). Semi-Supervised Classification with Graph Convolutional Networks. In International Conference on Learning Representations (ICLR). https://arxiv.org/abs/1609.02907 Introduced GCNs, foundational for node representation and message passing in graphs.
2. Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., & Bengio, Y. (2018). Graph Attention Networks. In International Conference on Learning Representations (ICLR). https://arxiv.org/abs/1710.10903 Proposed attention mechanisms for GNNs, improving flexibility and interpretability.
3. Scarselli, F., Gori, M., Tsoi, A. C., Hagenbuchner, M., & Monfardini, G. (2009). The Graph Neural Network Model. In IEEE Transactions on Neural Networks. DOI: 10.1109/TNN.2008.2005605 One of the earliest formalizations of GNNs.
4. Choi, T. Y., Dooley, K. J., & Rungtusanatham, M. (2001). Supply networks and complex adaptive systems: control versus emergence. In Journal of Operations Management, 19(3), 351–366. Discusses supply chains as complex networks, foundational for later graph-based modeling.
5. Nair, A., & Vidal, J. M. (2011). Supply network topology and robustness against disruptions – an investigation using multi-agent model. In International Journal of Production Research, 49(5), 1391–1404. DOI: 10.1080/00207543.2010.536902 Examines how disruption propagates across supply chain networks.
6. Thadakamalla, H. P., et al. (2004). Survivability of multiagent-based supply networks: A topological perspective.
7. In IEEE Intelligent Systems, 19(5), 24–31. DOI: 10.1109/MIS.2004.42 Risk propagation in networks studied through topology—early support for graph-based methods.
8. Pathak, S. D., Day, J. M., Nair, A., Sawaya, W. J., & Kristal, M. M. (2007). Complexity and adaptivity in supply networks: Building supply network theory using a complex adaptive systems perspective. In Decision Sciences, 38(4), 547–580. DOI: 10.1111/j.1540-5915.2007.00170.x Explores supply networks as dynamic systems—relevant for GNN-based modeling.
9. Bellamy, M. A., Ghosh, S., & Hora, M. (2014). The influence of supply network structure on firm innovation.
10.In Journal of Operations Management, 32(6), 357–373. DOI: 10.1016/j.jom.2014.09.001 Leverages network structures to study influence and flow—core concepts in GNNs.
11.Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., & Yu, P. S. (2019). A comprehensive survey on graph neural networks. In IEEE Transactions on Neural Networks and Learning Systems, 32(1), 4–24. DOI: 10.1109/TNNLS.2020.2978386
Although survey was published in early 2020, it only includes works until 2019. Still valid as pre-2020 source of methods.
12.Zhou, J., Cui, G., Zhang, Z., Yang, C., Liu, Z., Wang, L., & Sun, M. (2018). Graph Neural Networks: A Review of Methods and Applications. arXiv:1812.08434 https://arxiv.org/abs/1812.08434 Reviews early applications of GNNs in social, citation, and recommendation networks—methodologies transferable to supply chains.
