arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2106.03058cs.DS
收藏

近似图传播

Approximate Graph Propagation

Hanzhi Wang, Mingguo He, Zhewei Wei, Sibo Wang, Ye Yuan, Xiaoyong Du, Ji-Rong Wen

摘要:

节点邻近查询的有效计算,如转移概率,个性化PageRank和Katz在各种图挖掘和学习任务中具有根本的重要性。特别是,最近的几项工作利用快速节点邻近度计算来提高图神经网络(GNN)的可扩展性。然而,之前关于邻近计算和GNN特征传播的研究都是基于个案的,每篇论文都专注于特定的邻近度量。在本文中,我们提出了近似图传播(AGP),一个统一的随机算法,计算各种邻近查询和GNN特征传播,包括转移概率,个性化PageRank,热核PageRank,Katz,SGC,GDC和APPNP。我们的算法提供了一个理论上有界的错误保证,并运行在几乎最佳的时间复杂度。我们进行了广泛的实验研究,以证明AGP的有效性,在两个具体的应用程序:本地聚类与热核PageRank和节点分类与GNNs。最值得注意的是,我们提出了一个十亿边图Papers100M的实证研究,这是迄今为止最大的公开可用的GNN数据集。结果表明,AGP可以显着提高各种现有的GNN模型的可扩展性,而不牺牲预测精度。

英文摘要:

Efficient computation of node proximity queries such as transition probabilities, Personalized PageRank, and Katz are of fundamental importance in various graph mining and learning tasks. In particular, several recent works leverage fast node proximity computation to improve the scalability of Graph Neural Networks (GNN). However, prior studies on proximity computation and GNN feature propagation are on a case-by-case basis, with each paper focusing on a particular proximity measure. In this paper, we propose Approximate Graph Propagation (AGP), a unified randomized algorithm that computes various proximity queries and GNN feature propagation, including transition probabilities, Personalized PageRank, heat kernel PageRank, Katz, SGC, GDC, and APPNP. Our algorithm provides a theoretical bounded error guarantee and runs in almost optimal time complexity. We conduct an extensive experimental study to demonstrate AGP's effectiveness in two concrete applications: local clustering with heat kernel PageRank and node classification with GNNs. Most notably, we present an empirical study on a billion-edge graph Papers100M, the largest publicly available GNN dataset so far. The results show that AGP can significantly improve various existing GNN models' scalability without sacrificing prediction accuracy.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑