Local2Global: Unsupervised multi-view deep graph representation learning with Nearest Neighbor Constraint

作者：

Highlights：

• Fusing all features of single-view graph into a multi-view global graph feature.

• Providing a new Nearest Neighbor Constraint Variational Graph Auto-Encoder.

• Proposing a Mutli-view Deep Graph Representation Learning (MDGRL) framework.

• Proving that MDGRL framework outperforms other benchmark methods in experiments.

摘要

•Fusing all features of single-view graph into a multi-view global graph feature.•Providing a new Nearest Neighbor Constraint Variational Graph Auto-Encoder.•Proposing a Mutli-view Deep Graph Representation Learning (MDGRL) framework.•Proving that MDGRL framework outperforms other benchmark methods in experiments.

论文关键词：Multi-view,Deep graph representation learning,Variational Graph Auto-Encoder (VGAE),Nearest Neighbor Constraint (NNC)

论文评审过程：Received 20 December 2020, Revised 19 August 2021, Accepted 22 August 2021, Available online 25 August 2021, Version of Record 3 September 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.107439