Blockchain Analytics and Graph Neural Network Based Supplier Risk Detection for International Procurement Decision Systems

Abstract: This paper presents a blockchain analytics and graph neural network based supplier risk detection framework for international procurement decision systems. The study addresses persistent weaknesses in conventional supplier risk assessment models, including limited transaction traceability, fragmented procurement records, delayed fraud detection, and poor modelling of hidden supplier relationships across global supply networks. A novel Blockchain Graph Attention Risk Network BGARN is proposed to transform blockchain procurement transactions, supplier certifications, shipment events, smart contract records, payment histories, and cross-border compliance signals into a heterogeneous supplier risk graph. The model combines graph attention networks, temporal graph convolution, blockchain anomaly scoring, and supplier relationship embeddings to detect high-risk suppliers, collusive procurement clusters, delayed delivery patterns, contract non-compliance, and financial irregularities. The proposed BGARN model is compared with existing machine learning and deep learning algorithms, including Random Forest, XGBoost, Support Vector Machine, Multilayer Perceptron, Graph Convolutional Network, GraphSAGE, and standard Graph Attention Network. Experimental graphs are used to visualize supplier risk propagation, anomaly clusters, model accuracy, precision-recall performance, and comparative detection efficiency. Results are expected to demonstrate that BGARN achieves superior supplier risk classification, higher fraud detection sensitivity, lower false positive rates, and stronger interpretability for procurement decision makers. The paper contributes a technically advanced decision-support architecture for international procurement by integrating immutable blockchain audit trails with adaptive graph-based risk intelligence. The framework supports transparent supplier evaluation, early warning detection, compliance monitoring, and data-driven procurement governance in complex international supply chains. Keywords: Blockchain Analytics; Graph Neural Network; Supplier Risk Detection; International Procurement; Decision Systems. Title: Blockchain Analytics and Graph Neural Network Based Supplier Risk Detection for International Procurement Decision Systems Author: Bright Amankwah, Joy Onma Enyejo International Journal of Recent Research in Commerce Economics and Management (IJRRCEM) ISSN 2349-7807 Vol. 13, Issue 2, April 2026 - June 2026 Page No: 173-194 Paper Publications Website: www.paperpublications.org Published Date: 16-June-2026 DOI: https://doi.org/10.5281/zenodo.20715516 Paper Download Link (Source) https://www.paperpublications.org/upload/book/Blockchain%20Analytics%20and%20Graph%20Neural-16062026-2.pdf