Simple Multigraph Convolution Networks

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wu, Danyang, Shen, Xinjie, Lu, Jitao, Xu, Jin, Nie, Feiping
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911791550824448
author Wu, Danyang
Shen, Xinjie
Lu, Jitao
Xu, Jin
Nie, Feiping
author_facet Wu, Danyang
Shen, Xinjie
Lu, Jitao
Xu, Jin
Nie, Feiping
contents Existing multigraph convolution methods either ignore the cross-view interaction among multiple graphs, or induce extremely high computational cost due to standard cross-view polynomial operators. To alleviate this problem, this paper proposes a Simple MultiGraph Convolution Networks (SMGCN) which first extracts consistent cross-view topology from multigraphs including edge-level and subgraph-level topology, then performs polynomial expansion based on raw multigraphs and consistent topologies. In theory, SMGCN utilizes the consistent topologies in polynomial expansion rather than standard cross-view polynomial expansion, which performs credible cross-view spatial message-passing, follows the spectral convolution paradigm, and effectively reduces the complexity of standard polynomial expansion. In the simulations, experimental results demonstrate that SMGCN achieves state-of-the-art performance on ACM and DBLP multigraph benchmark datasets. Our codes are available at https://github.com/frinkleko/SMGCN.
format Preprint
id arxiv_https___arxiv_org_abs_2403_05014
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Simple Multigraph Convolution Networks
Wu, Danyang
Shen, Xinjie
Lu, Jitao
Xu, Jin
Nie, Feiping
Machine Learning
Artificial Intelligence
Existing multigraph convolution methods either ignore the cross-view interaction among multiple graphs, or induce extremely high computational cost due to standard cross-view polynomial operators. To alleviate this problem, this paper proposes a Simple MultiGraph Convolution Networks (SMGCN) which first extracts consistent cross-view topology from multigraphs including edge-level and subgraph-level topology, then performs polynomial expansion based on raw multigraphs and consistent topologies. In theory, SMGCN utilizes the consistent topologies in polynomial expansion rather than standard cross-view polynomial expansion, which performs credible cross-view spatial message-passing, follows the spectral convolution paradigm, and effectively reduces the complexity of standard polynomial expansion. In the simulations, experimental results demonstrate that SMGCN achieves state-of-the-art performance on ACM and DBLP multigraph benchmark datasets. Our codes are available at https://github.com/frinkleko/SMGCN.
title Simple Multigraph Convolution Networks
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2403.05014