TransGCN:Coupling Transformation Assumptions with Graph Convolutional Networks for Link Prediction

TL;DR

TransGCN introduces a unified graph convolutional network framework for link prediction in knowledge graphs, integrating relation and entity embedding learning with transformation assumptions, outperforming existing methods.

Contribution

It proposes a novel TransGCN framework that models heterogeneous relations as transformations and learns relation and entity embeddings simultaneously, reducing parameters and improving performance.

Findings

Outperforms state-of-the-art on FB15K-237 and WN18RR datasets.

Incorporates transformation assumptions like translation and rotation.

Learns relation and entity embeddings jointly during convolution.

Abstract

Link prediction is an important and frequently studied task that contributes to an understanding of the structure of knowledge graphs (KGs) in statistical relational learning. Inspired by the success of graph convolutional networks (GCN) in modeling graph data, we propose a unified GCN framework, named TransGCN, to address this task, in which relation and entity embeddings are learned simultaneously. To handle heterogeneous relations in KGs, we introduce a novel way of representing heterogeneous neighborhood by introducing transformation assumptions on the relationship between the subject, the relation, and the object of a triple. Specifically, a relation is treated as a transformation operator transforming a head entity to a tail entity. Both translation assumption in TransE and rotation assumption in RotatE are explored in our framework. Additionally, instead of only learning entity embeddings in the convolution-based encoder while learning relation embeddings in the decoder as done by the state-of-art models, e.g., R-GCN, the TransGCN framework trains relation embeddings and entity embeddings simultaneously during the graph convolution operation, thus having fewer parameters compared with R-GCN. Experiments show that our models outperform the-state-of-arts methods on both FB15K-237 and WN18RR.