Topology and Content Co-Alignment Graph Convolutional Learning
Topology and Content Co-Alignment Graph Convolutional Learning
复制标题
DOI：
10.1109/tnnls.2021.3084125
复制
发表时间：
2020-03
影响因子：
10.4
通讯作者：
Min Shi;Yufei Tang;Xingquan Zhu
中科院分区：
文献类型：
--
作者：
Min Shi;Yufei Tang;Xingquan Zhu
在传统的图神经网络（GNN）中，图卷积学习是通过拓扑驱动的递归节点内容聚合进行网络表示学习。实际上，网络拓扑和节点内容各自提供独特且重要的信息，并且由于节点之间的噪声、不相关性或缺失链接，它们并不总是一致的。由于不正确的消息在整个网络上的传播，未对齐邻域之间的纯拓扑驱动的特征聚合方法可能会恶化结构内容一致性差的节点的学习效果。或者，在本简介中，我们提倡共同对齐图卷积学习（CoGL）范式，通过对齐拓扑和内容网络来最大限度地提高一致性。我们的主题是强制拓扑网络的学习与内容网络保持一致，同时优化内容网络以符合优化表示学习的拓扑。给定一个网络，CoGL 首先根据节点特征重建内容网络，然后通过统一的优化目标将内容网络和原始网络共同对齐：1）最小化内容损失； 2）最小化分类损失； 3）最小化对抗性损失。六个基准测试的实验表明，与现有最先进的 GNN 模型相比，CoGL 实现了可比甚至更好的性能。
In traditional graph neural networks (GNNs), graph convolutional learning is carried out through topology-driven recursive node content aggregation for network representation learning. In reality, network topology and node content each provide unique and important information, and they are not always consistent because of noise, irrelevance, or missing links between nodes. A pure topology-driven feature aggregation approach between unaligned neighborhoods may deteriorate learning from nodes with poor structure-content consistency, due to the propagation of incorrect messages over the whole network. Alternatively, in this brief, we advocate a co-alignment graph convolutional learning (CoGL) paradigm, by aligning topology and content networks to maximize consistency. Our theme is to enforce the learning from the topology network to be consistent with the content network while simultaneously optimizing the content network to comply with the topology for optimized representation learning. Given a network, CoGL first reconstructs a content network from node features then co-aligns the content network and the original network through a unified optimization goal with: 1) minimized content loss; 2) minimized classification loss; and 3) minimized adversarial loss. Experiments on six benchmarks demonstrate that CoGL achieves comparable and even better performance compared with existing state-of-the-art GNN models.