Attributed Multi-order Graph Convolutional Network for Heterogeneous Graphs

TL;DR

This paper introduces AMOGCN, a novel heterogeneous graph neural network that automatically learns multi-hop relations through adaptive multi-order adjacency matrices, enhancing node classification performance.

Contribution

The paper proposes an automatic meta-path learning method using multi-order adjacency matrices fused by semantic supervision, improving heterogeneous graph embedding quality.

Findings

AMOGCN outperforms state-of-the-art methods in semi-supervised classification.

The adaptive fusion of multi-order adjacency matrices enhances embedding discriminability.

Semantic supervision effectively guides the learning of multi-hop relations.

Abstract

Heterogeneous graph neural networks aim to discover discriminative node embeddings and relations from multi-relational networks.One challenge of heterogeneous graph learning is the design of learnable meta-paths, which significantly influences the quality of learned embeddings.Thus, in this paper, we propose an Attributed Multi-Order Graph Convolutional Network (AMOGCN), which automatically studies meta-paths containing multi-hop neighbors from an adaptive aggregation of multi-order adjacency matrices. The proposed model first builds different orders of adjacency matrices from manually designed node connections. After that, an intact multi-order adjacency matrix is attached from the automatic fusion of various orders of adjacency matrices. This process is supervised by the node semantic information, which is extracted from the node homophily evaluated by attributes. Eventually, we utilize a one-layer simplifying graph convolutional network with the learned multi-order adjacency matrix, which is equivalent to the cross-hop node information propagation with multi-layer graph neural networks. Substantial experiments reveal that AMOGCN gains superior semi-supervised classification performance compared with state-of-the-art competitors.