Hierarchical Contrastive Learning Enhanced Heterogeneous Graph Neural Network

TL;DR

HeCo++ conducts hierarchical contrastive learning, including cross-view and intra-view contrasts, which aims to enhance the mining of respective structures and shows superior performance over the state-of-the-arts.

Abstract

Heterogeneous graph neural networks (HGNNs) as an emerging technique have shown superior capacity of dealing with heterogeneous information network (HIN). However, most HGNNs follow a semi-supervised learning manner, which notably limits their wide use in reality since labels are usually scarce in real applications. Recently, contrastive learning, a self-supervised method, becomes one of the most exciting learning paradigms and shows great potential when there are no labels. In this paper, we study the problem of self-supervised HGNNs and propose a novel co-contrastive learning mechanism for HGNNs, named HeCo. Different from traditional contrastive learning which only focuses on contrasting positive and negative samples, HeCo employs cross-view contrastive mechanism. Specifically, two views of a HIN (network schema and meta-path views) are proposed to learn node embeddings, so as to capture both of local and high-order structures simultaneously. Then the cross-view contrastive learning, as well as a view mask mechanism, is proposed, which is able to extract the positive and negative embeddings from two views. This enables the two views to collaboratively supervise each other and finally learn high-level node embeddings. Moreover, to further boost the performance of HeCo, two additional methods are designed to generate harder negative samples with high quality. Besides the invariant factors, view-specific factors complementally provide the diverse structure information between different nodes, which also should be contained into the final embeddings. Therefore, we need to further explore each view independently and propose a modified model, called HeCo++. Specifically, HeCo++ conducts hierarchical contrastive learning, including cross-view and intra-view contrasts, which aims to enhance the mining of respective structures.

Figures (14)

• Figure 1: A toy example of HIN (ACM) and relative illustrations of meta-path and network schema.
• Figure 2: The overall architecture of our proposed HeCo.
• Figure 3: A schematic diagram of view mask mechanism.
• Figure 4: The partial architecture of our extended HeCo++. (T: target, P: positive, N: negative)
• Figure 5: Visualization of the learned node embedding on ACM. The Silhouette scores for (a) (b) (c) (d) (e) are 0.0292, 0.1862, 0.3015, 0.3642 and 0.3885, respectively.