A Path Recommendation Method Based on the Siamese Graph Convolutional Network for the Holographic Counterpart of Consumer Electronics Logistics of the Internet of Things

The proliferation of Internet of Things (IoT) devices brings unprecedented challenges and opportunities to the consumer electronics (CE) logistics industry. To meet its highly dynamic and predictive path optimization needs, this study introduces the cutting-edge concept of "holographic counterpart"—a dynamic virtual replica that synchronizes in real time with the physical CE logistics network. This study proposes an innovative deep learning framework, the Cross-domain Bilinear Attention Network-Siamese Graph Convolutional Network (CDBAN-SGCN), as the intelligent engine driving this holographic counterpart. The framework first utilizes data mining (DM) technology to construct a graph-structured representation of the CE network from IoT data streams. Finally, a Siamese Network is utilized to measure the similarity between paths, combined with a CDBAN, to construct a path recommendation model supported by the CDBAN-SGCN algorithm for intelligent path recommendations. The results show that the proposed CDBAN-SGCN algorithm achieves accuracies of 95.17% and 96.47%, precisions of 88.49% and 82.91%, recalls of 78.98% and 77.40%, and F1 scores of 83.98% and 81.35% on the Amazon Fresh and Alibaba datasets. These results indicate that the CDBAN-SGCN algorithm outperforms the GCN algorithm in supply chain path recommendations. Therefore, this study successfully develops and verifies a novel path recommendation model. This model can effectively improve transportation efficiency, reduce logistics costs, and enhance the resilience of the CE supply chain. Thus, the model can provide a powerful and empirically tested solution for intelligent logistics optimization in CE.