Department of Electrical and Computer Engineering, University of Pittsburgh, 3700 O'Hara St, Pittsburgh, PA, 15260, USA.
Intel Labs, 2111 NE 25th Ave, Hillsboro, OR, 97124, USA.
Neural Netw. 2021 Nov;143:669-677. doi: 10.1016/j.neunet.2021.07.028. Epub 2021 Jul 29.
Recent years have witnessed the emergence and flourishing of hierarchical graph pooling neural networks (HGPNNs) which are effective graph representation learning approaches for graph level tasks such as graph classification. However, current HGPNNs do not take full advantage of the graph's intrinsic structures (e.g., community structure). Moreover, the pooling operations in existing HGPNNs are difficult to be interpreted. In this paper, we propose a new interpretable graph pooling framework - CommPOOL, that can capture and preserve the hierarchical community structure of graphs in the graph representation learning process. Specifically, the proposed community pooling mechanism in CommPOOL utilizes an unsupervised approach for capturing the inherent community structure of graphs in an interpretable manner. CommPOOL is a general and flexible framework for hierarchical graph representation learning that can further facilitate various graph-level tasks. Evaluations on five public benchmark datasets and one synthetic dataset demonstrate the superior performance of CommPOOL in graph representation learning for graph classification compared to the state-of-the-art baseline methods, and its effectiveness in capturing and preserving the community structure of graphs.
近年来，层次图池化神经网络（HGPNNs）的出现和发展见证了其作为图级任务（例如图分类）的有效图表示学习方法的出现和发展。然而，目前的 HGPNNs 并没有充分利用图的内在结构（例如社区结构）。此外，现有 HGPNNs 中的池化操作难以解释。在本文中，我们提出了一种新的可解释图池化框架——CommPOOL，该框架可以在图表示学习过程中捕获和保留图的层次社区结构。具体来说，CommPOOL 中的提出的社区池化机制利用了一种无监督的方法，可以以可解释的方式捕获图的内在社区结构。CommPOOL 是一种通用且灵活的层次图表示学习框架，可进一步促进各种图级任务。在五个公共基准数据集和一个合成数据集上的评估表明，与最先进的基线方法相比，CommPOOL 在图分类等图表示学习任务中的性能更优，并且在捕获和保留图的社区结构方面也更有效。
