The field of graph neural networks and transportation systems is rapidly evolving, with a focus on developing innovative methods for predicting mobility patterns, optimizing supply chain transportation, and improving link prediction in graphs. Researchers are exploring new architectures, such as integrated graph-based neural networks and hybrid models combining convolutional and graph neural networks, to capture complex patterns and relationships in data. Notable papers in this area include the proposal of SP4LP, a novel framework that combines GNN-based node encodings with sequence modeling over shortest paths, and Context Pooling, a methodology that applies graph pooling in knowledge graphs to enhance GNN-based models' efficacy for link predictions. These advancements have the potential to significantly impact various applications, from transportation planning and infrastructure management torecommendation systems and social network analysis.
Advances in Graph Neural Networks and Transportation Systems
Sources
Dynamic Campus Origin-Destination Mobility Prediction using Graph Convolutional Neural Network on WiFi Logs
Predicting Graph Structure via Adapted Flux Balance Analysis
GNNs Meet Sequence Models Along the Shortest-Path: an Expressive Method for Link Prediction
Supply Chain Optimization via Generative Simulation and Iterative Decision Policies
GNN-CNN: An Efficient Hybrid Model of Convolutional and Graph Neural Networks for Text Representation
Context Pooling: Query-specific Graph Pooling for Generic Inductive Link Prediction in Knowledge Graphs
Beyond Connectivity: Higher-Order Network Framework for Capturing Memory-Driven Mobility Dynamics