Abstract:Knowledge Graph Embedding (KGE) aims to learn representations of entities and relations of knowledge graph (KG). Recently Graph Neural Networks (GNNs) have gained great success on KGE, but for the reason behind it, most views simply attribute to the well learning of knowledge graph structure, which still remains a limited understanding of the internal mechanism. In this work, we first study a fundamental problem, i.e., what are the important factors for GNNs to help KGE. To investigate this problem, we discuss the core idea of current GNN models for KG, and propose a new assumption of relational homophily that connected nodes possess similar features after relation's transforming, to explain why aggregating neighbors with relation can help KGE. Based on the model and empirical analyses, we then introduce a novel data-centric framework for applying GNNs to KGE called KSG-GNN . In KSG-GNN, we construct a new graph structure from KG named Knowledge Similarity Graph (KSG), where each node connects with its similar nodes as neighbors, and then we apply GNNs on this graph to perform KGE. Instead of following the relational homophily assumption in KG, KSG aligns with homogeneous graphs that can directly satisfy homophily assumption. Hence, any GNN developed on homogeneous graphs like GCN, GAT, GraphSAGE, etc., can be applied out-of-the-box as KSG-GNN without modification, which provides a more general and effective GNN paradigm. Finally, we conduct extensive experiments on two benchmark datasets, i.e., FB15k-237 and WN18RR, demonstrating the superior performance of KSG-GNN over multiple strong baselines. The source code is available at https://github.com/advancer99/WWWJ-KGE.

Distributed Representations of Entities in Open-World Knowledge Graphs

Distributed representations of entities in open-world knowledge graphs

Multi-level Disentanglement Graph Neural Network

An Entity Alignment Method Based on Graph Attention Network with Pre-classification

A data-centric framework of improving graph neural networks for knowledge graph embedding

Graph neural networks meet with distributed graph partitioners and reconciliations

Distributed Representation Learning for Knowledge Graphs with Entity Descriptions

GoGNN: Graph of Graphs Neural Network for Predicting Structured Entity Interactions

Entity alignment via graph neural networks: a component-level study

Knowledge Graph Alignment Network with Gated Multi-Hop Neighborhood Aggregation

A cross-linguistic entity alignment method based on graph convolutional neural network and graph attention network

Iteratively Learning Representations for Unseen Entities with Inter-Rule Correlations

Powerful Graph of Graphs Neural Network for Structured Entity Analysis

SubGraph Networks based Entity Alignment for Cross-lingual Knowledge Graph

Multi-Relational Graph Representation Learning with Bayesian Gaussian Process Network.

Synergistic Signals: Exploiting Co-Engagement and Semantic Links via Graph Neural Networks

Knowledge Graph Embedding with Entity Neighbors and Deep Memory Network

DPGNN: Dual-perception graph neural network for representation learning

Learning knowledge graph embedding with a dual-attention embedding network

Improving Entity Linking with Graph Networks

Online Adversarial Knowledge Distillation for Graph Neural Networks