Incorporating Siamese Network Structure into Graph Neural Network

Abstract Siamese network plays an important role in many artificial intelligence domains, but there requires more exploration of applying structure to graph neural network. This paper proposes a novel framework that incorporates into Graph Neural Network (Siam-GNN). We use DropEdge as augmentation technique generate new graphs. Besides, the strategy constructing network’s paired inputs is also studied our work. Notably, stopping gradient backpropagation one side Siam-GNN factor affecting performance model. equip some networks with and evaluate these Siam-GNNs on several standard semi-supervised node classification datasets achieve surprising improvement almost every original