TrustSGCN: Learning Trustworthiness on Edge Signs for Effective Signed Graph Convolutional Networks

Abstract

The problem of signed network embedding (SNE) aims to represent nodes in a given signed network as low-dimensional vectors. While several SNE methods based on graph convolutional networks (GCN) have been proposed, we point out that they significantly rely on the assumption that the decades-old balance theory always holds in the real world. To address this limitation, we propose a novel GCN-based SNE approach, named as TrustSGCN, which measures the trustworthiness on edge signs for high-order relationships inferred by balance theory and corrects incorrect embedding propagation based on the trustworthiness. The experiments on four real-world signed network datasets demonstrate that TrustSGCN consistently outperforms five state-of-the-art GCN-based SNE methods. The code is available at https://github.com/kmj0792/TrustSGCN.