Graph Learning Network: A Structure Learning Algorithm
This addresses the problem of handling dynamic or learned graph structures in tasks like community detection and node classification for researchers and practitioners using GNNs, but it appears incremental as it builds on existing GNN methods.
The paper tackles the limitation of static relationships in graph neural networks (GNNs) by proposing the Graph Learning Network (GLN), a recursive algorithm that learns node embeddings and predicts graph structures, resulting in enhanced predictions and embeddings.
Recently, graph neural networks (GNNs) have proved to be suitable in tasks on unstructured data. Particularly in tasks as community detection, node classification, and link prediction. However, most GNN models still operate with static relationships. We propose the Graph Learning Network (GLN), a simple yet effective process to learn node embeddings and structure prediction functions. Our model uses graph convolutions to propose expected node features, and predict the best structure based on them. We repeat these steps recursively to enhance the prediction and the embeddings.