典型文献
NGAT:attention in breadth and depth exploration for semi-supervised graph representation learning
文献摘要：
Recently, graph neural networks (GNNs) have achieved remarkable performance in representation learning on graph-structured data. However, as the number of network layers increases, GNNs based on the neighborhood aggregation strategy deteriorate due to the problem of oversmoothing, which is the major bottleneck for applying GNNs to real-world graphs. Many efforts have been made to improve the process of feature information aggregation from directly connected nodes, i.e., breadth exploration. However, these models perform the best only in the case of three or fewer layers, and the performance drops rapidly for deep layers. To alleviate oversmoothing, we propose a nested graph attention network (NGAT), which can work in a semi-supervised manner. In addition to breadth exploration, a k-layer NGAT uses a layer-wise aggregation strategy guided by the attention mechanism to selectively leverage feature information from the kth-order neighborhood, i.e., depth exploration. Even with a 10-layer or deeper architecture, NGAT can balance the need for preserving the locality (including root node features and the local structure) and aggregating the information from a large neighborhood. In a number of experiments on standard node classification tasks, NGAT outperforms other novel models and achieves state-of-the-art performance.
文献关键词：
中图分类号：
作者姓名：
Jianke HU;Yin ZHANG
作者机构：
College of Computer Science and Technology,Zhejiang University,Hangzhou 310027,China
文献出处：
引用格式：
[1]Jianke HU;Yin ZHANG-.NGAT:attention in breadth and depth exploration for semi-supervised graph representation learning)[J].信息与电子工程前沿（英文）,2022(03):409-421
A类：
NGAT,oversmoothing,kth
B类：
attention,breadth,depth,exploration,semi,supervised,representation,learning,Recently,neural,networks,GNNs,have,achieved,remarkable,performance,structured,data,However,number,layers,increases,neighborhood,aggregation,strategy,deteriorate,due,problem,which,major,bottleneck,applying,real,world,graphs,Many,efforts,been,made,improve,process,information,from,directly,connected,nodes,these,models,best,only,case,three,fewer,drops,rapidly,To,alleviate,propose,nested,can,manner,In,addition,uses,wise,guided,by,mechanism,selectively,leverage,order,Even,deeper,architecture,balance,need,preserving,locality,including,root,features,aggregating,large,experiments,standard,classification,tasks,outperforms,other,novel,achieves,state,art
AB值：
0.500901
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
