Wasserstein Graph Neural Networks for Graphs with Missing Attributes
This work provides an incremental improvement for graph neural networks dealing with incomplete attribute information, which is a common problem in real-world graph data.
The paper addresses the problem of missing node attributes in graph neural networks by proposing Wasserstein Graph Neural Network (WGNN). It represents nodes as low-dimensional distributions and aggregates distributional information in the Wasserstein space, demonstrating superior performance in node classification and matrix completion tasks on various datasets.
Missing node attributes is a common problem in real-world graphs. Graph neural networks have been demonstrated power in graph representation learning while their performance is affected by the completeness of graph information. Most of them are not specified for missing-attribute graphs and fail to leverage incomplete attribute information effectively. In this paper, we propose an innovative node representation learning framework, Wasserstein Graph Neural Network (WGNN), to mitigate the problem. To make the most of limited observed attribute information and capture the uncertainty caused by missing values, we express nodes as low-dimensional distributions derived from the decomposition of the attribute matrix. Furthermore, we strengthen the expressiveness of representations by developing a novel message passing schema that aggregates distributional information from neighbors in the Wasserstein space. We test WGNN in node classification tasks under two missing-attribute cases on both synthetic and real-world datasets. In addition, we find WGNN suitable to recover missing values and adapt them to tackle matrix completion problems with graphs of users and items. Experimental results on both tasks demonstrate the superiority of our method.