Complex graph convolutional network for link prediction in knowledge graphs

Knowledge graph (KG) embedding models map nodes and edges to fixed-length vectors and obtain the similarity of nodes as the output of a scoring function to predict missing links between nodes. KG embedding methods based on graph convolutional networks (GCNs) have recently gained significant attentio...

Full description

Saved in:
Bibliographic Details
Published in:Expert systems with applications 2022-08, Vol.200, p.116796, Article 116796
Main Authors: Zeb, Adnan, Saif, Summaya, Chen, Junde, Haq, Anwar Ul, Gong, Zhiguo, Zhang, Defu
Format: Article
Language:English
Subjects:
Artificial neural networks
Complex embeddings
Embedding
Graph convolutional network
Graph theory
Graphs
Knowledge graph
Knowledge representation
Link prediction
Network analysis
Nodes
Tensor decomposition
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Knowledge graph (KG) embedding models map nodes and edges to fixed-length vectors and obtain the similarity of nodes as the output of a scoring function to predict missing links between nodes. KG embedding methods based on graph convolutional networks (GCNs) have recently gained significant attention due to their ability to add information of neighboring nodes into the nodes’ embeddings. However, existing GCNs are primarily based on real-valued embeddings, which have high distortion, particularly when modeling graphs with varying geometric structures. In this paper, we propose complex graph convolutional network (ComplexGCN), a novel extension of the standard GCNs in complex space to combine the expressiveness of complex geometry with GCNs for improving the representation quality of KG components. The proposed ComplexGCN comprises a set of complex graph convolutional layers and a complex scoring function based on PARATUCK2 decomposition: the former includes information of neighboring nodes into the nodes’ embeddings, while the latter leverages these embeddings to predict new links between nodes. The proposed model demonstrates enhanced performance compared to existing methods on the two recent standard link prediction datasets. •Real embeddings usually suffer from high distortion.•An extension of graph convolution network is introduced with complex embeddings.•A novel complex scoring function based on PARATUCK2 decomposition is introduced.•The results show that complex embeddings lead to performance improvement.
ISSN:0957-4174
1873-6793
DOI:10.1016/j.eswa.2022.116796