[论文解读] Graph-MVP: Multi-View Prototypical Contrastive Learning for Multiplex Graphs

Baoyu Jing, Yuejia Xiang+3 more|arXiv (Cornell University)|Sep 8, 2021

Advanced Graph Neural Networks参考文献 63被引用 10

一句话总结

本文提出 Graph-MVP，一种用于多层图的多视图原型对比学习框架，解决了对比学习中的语义错误问题，并利用多种关系视图。通过结合图原型对比学习与基于硬聚类和软聚类策略的视图对齐，Graph-MVP 提升了节点表示学习效果，在真实世界数据集的下游任务中实现了最先进性能。

ABSTRACT

Contrastive Learning (CL) is one of the most popular self-supervised learning frameworks for graph representation learning, which trains a Graph Neural Network (GNN) by discriminating positive and negative node pairs. However, there are two challenges for CL on graphs. On the one hand, traditional CL methods will unavoidably introduce semantic errors since they will treat some semantically similar nodes as negative pairs. On the other hand, most of the existing CL methods ignore the multiplexity nature of the real-world graphs, where nodes are connected by various relations and each relation represents a view of the graph. To address these challenges, we propose a novel Graph Multi-View Prototypical (Graph-MVP) framework to extract node embeddings on multiplex graphs. Firstly, we introduce a Graph Prototypical Contrastive Learning (Graph-PCL) framework to capture both node-level and semantic-level information for each view of multiplex graphs. Graph-PCL captures the node-level information by a simple yet effective data transformation technique. It captures the semantic-level information by an Expectation-Maximization (EM) algorithm, which alternatively performs clustering over node embeddings and parameter updating for GNN. Next, we introduce Graph-MVP based on Graph-PCL to jointly model different views of the multiplex graphs. Our key insight behind Graph-MVP is that different view-specific embeddings of the same node should have similar underlying semantic, based on which we propose two versions of Graph-MVP: Graph-MVP_hard and Graph-MVP_soft to align embeddings across views. Finally, we evaluate the proposed Graph-PCL and Graph-MVP on a variety of real-world datasets and downstream tasks. The experimental results demonstrate the effectiveness of the proposed Graph-PCL and Graph-MVP frameworks.

研究动机与目标

• 解决对比学习中语义相似节点被错误地视为负样本对的问题。
• 通过将每种关系建模为独立视图，利用真实世界图中的多层结构，即节点通过多种关系连接。
• 通过联合建模多个视图，结合原型聚类与跨视图对齐，提升节点表示学习效果。
• 开发一个统一框架，以自监督方式捕捉视图间节点级与语义级信息。

提出的方法

• 提出 Graph-PCL，一种原型对比学习框架，通过数据增强实现节点级对比，利用期望最大化（EM）算法捕捉语义级结构。
• 采用基于 EM 的聚类过程，交替执行节点嵌入聚类与 GNN 参数更新，以优化语义表示。
• 在 Graph-MVP 中应用多视图学习策略，联合建模多层图中同一节点的不同关系视图。
• 提出两种变体——Graph-MVP_hard 和 Graph-MVP_soft，分别采用硬聚类与软聚类策略，基于共享语义内容对齐视图特定嵌入。
• 结合实例级与原型级对比学习，增强节点表示中对局部与全局结构的感知能力。
• 通过鼓励一致的聚类分配并最小化视图间表示的差异，实现跨视图嵌入对齐。

实验结果

研究问题

• RQ1 原型对比学习能否通过捕捉潜在节点语义，减少图对比学习中的语义错误？
• RQ2 如何有效利用多层图中的多个关系视图来提升节点表示学习？
• RQ3 跨视图对齐对多关系图中学习到的节点嵌入质量有何影响？
• RQ4 硬聚类与软聚类策略对多视图图表示学习性能有何影响？

主要发现

• Graph-MVP 在多种真实世界多层图数据集的下游节点分类与聚类任务中均实现了最先进性能。
• Graph-PCL 组件通过基于 EM 的聚类与原型学习显著提升了表示质量，有效减少了语义错误。
• Graph-MVP_hard 与 Graph-MVP_soft 均优于基线对比方法，其中软对齐变体在具有复杂关系结构的数据集中表现更优。
• 消融实验表明，视图级对比学习与跨视图对齐对实现最佳性能均至关重要。