Abstract:Graph Neural Networks (GNNs) have been widely used for the representation learning of various structured graph data, typically through message passing among nodes by aggregating their neighborhood information via different operations. While promising, most existing GNNs oversimplify the complexity and diversity of the edges in the graph and thus are inefficient to cope with ubiquitous heterogeneous graphs, which are typically in the form of multi-relational graph representations. In this article, we propose RioGNN , a novel Reinforced, recursive, and flexible neighborhood selection guided multi-relational Graph Neural Network architecture, to navigate complexity of neural network structures whilst maintaining relation-dependent representations. We first construct a multi-relational graph, according to the practical task, to reflect the heterogeneity of nodes, edges, attributes, and labels. To avoid the embedding over-assimilation among different types of nodes, we employ a label-aware neural similarity measure to ascertain the most similar neighbors based on node attributes. A reinforced relation-aware neighbor selection mechanism is developed to choose the most similar neighbors of a targeting node within a relation before aggregating all neighborhood information from different relations to obtain the eventual node embedding. Particularly, to improve the efficiency of neighbor selecting, we propose a new recursive and scalable reinforcement learning framework with estimable depth and width for different scales of multi-relational graphs. RioGNN can learn more discriminative node embedding with enhanced explainability due to the recognition of individual importance of each relation via the filtering threshold mechanism. Comprehensive experiments on real-world graph data and practical tasks demonstrate the advancements of effectiveness, efficiency, and the model explainability, as opposed to other comparative GNN models.

Dynamic Graph Neural Networks with Neighborhood Constraints

Discovering the Representation Bottleneck of Graph Neural Networks from Multi-order Interactions

Neighborhood Convolutional Graph Neural Network

Neighborhood Extended Dynamic Graph Neural Network.

Dynamic Spiking Framework for Graph Neural Networks

Deep Graph Neural Networks via Flexible Subgraph Aggregation

Instant Graph Neural Networks for Dynamic Graphs

Graph Elimination Networks

Graph Convolutional Networks with Adaptive Neighborhood Awareness

Graph neural networks via contrast between separation and aggregation for self and neighborhood

Graph Representation Learning on Noise and Sparse Labels.

Reinforced Neighborhood Selection Guided Multi-Relational Graph Neural Networks

Graph Representation Learning Via Simple Jumping Knowledge Networks

Local Clustering Graph Neural Networks

Efficient and Effective Implicit Dynamic Graph Neural Network

Two-level Graph Neural Network

Dynamic Spiking Graph Neural Networks.

Geodesic Graph Neural Network for Efficient Graph Representation Learning

Graph Structure Estimation Neural Networks

Dynamic Personalized Graph Neural Network with Linear Complexity for Multivariate Time Series Forecasting

Neighborhood Convolutional Network: A New Paradigm of Graph Neural Networks for Node Classification