AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning Tasks

Graph neural networks (GNNs), particularly graph convolutional networks (GCNs), have demonstrated remarkable success in modeling graph-structured data across diverse applications. A critical yet underexplored aspect of GCN design lies in graph representation normalization, where the choice of normal...

Full description

Saved in:
Bibliographic Details
Main Authors: Chuan Dai, Yajuan Wei, Hao Wang, Ying Liu, Zhijie Xu
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11059875/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849390183896580096
author Chuan Dai
Yajuan Wei
Hao Wang
Ying Liu
Zhijie Xu
author_facet Chuan Dai
Yajuan Wei
Hao Wang
Ying Liu
Zhijie Xu
author_sort Chuan Dai
collection DOAJ
description Graph neural networks (GNNs), particularly graph convolutional networks (GCNs), have demonstrated remarkable success in modeling graph-structured data across diverse applications. A critical yet underexplored aspect of GCN design lies in graph representation normalization, where the choice of normalization scheme significantly influences model performance on learning tasks. Despite its importance, existing research lacks systematic analysis on determining optimal task-specific normalization strategies. This study proposes AdaptedNorm, a principled framework that establishes connections between normalization schemes and GCN task objectives, while introducing three key contributions: 1) A novel random walk re-normalized transformation (RWRT) technique for graph representation normalization; 2) A task-aligned modeling paradigm demonstrating that symmetric re-normalized transformation (SRT) enhances node classification accuracy, while RWRT achieves superior performance in link prediction; 3) Integration of normalization schemes with GNN sparsification strategies, enabling effective model compression without sacrificing performance. Extensive experiments on benchmark datasets (Cora, Citeseer, and PubMed) confirm the generalizability of the findings. The implementation is publicly available at: <uri>https://www.github.com/ChuanDai/AdaptedNorm</uri>.
format Article
id doaj-art-d9f89752aa824f938b45c18f89909d0a
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-d9f89752aa824f938b45c18f89909d0a2025-08-20T03:41:44ZengIEEEIEEE Access2169-35362025-01-011311486511487910.1109/ACCESS.2025.358452311059875AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning TasksChuan Dai0https://orcid.org/0009-0008-5082-7141Yajuan Wei1Hao Wang2Ying Liu3Zhijie Xu4https://orcid.org/0000-0002-0524-5926School of Intelligent Science and Engineering, Xi&#x2019;an Peihua University, Xi&#x2019;an, ChinaSchool of Cyberspace Security, Xi&#x2019;an University of Posts and Telecommunications, Xi&#x2019;an, ChinaSchool of Economics and Management, Xi&#x2019;an University of Technology, Xi&#x2019;an, ChinaSchool of Communications and Information Engineering, Xi&#x2019;an University of Posts and Telecommunications, Xi&#x2019;an, ChinaSchool of Advanced Technology, Xi&#x2019;an Jiaotong-Liverpool University, Suzhou, ChinaGraph neural networks (GNNs), particularly graph convolutional networks (GCNs), have demonstrated remarkable success in modeling graph-structured data across diverse applications. A critical yet underexplored aspect of GCN design lies in graph representation normalization, where the choice of normalization scheme significantly influences model performance on learning tasks. Despite its importance, existing research lacks systematic analysis on determining optimal task-specific normalization strategies. This study proposes AdaptedNorm, a principled framework that establishes connections between normalization schemes and GCN task objectives, while introducing three key contributions: 1) A novel random walk re-normalized transformation (RWRT) technique for graph representation normalization; 2) A task-aligned modeling paradigm demonstrating that symmetric re-normalized transformation (SRT) enhances node classification accuracy, while RWRT achieves superior performance in link prediction; 3) Integration of normalization schemes with GNN sparsification strategies, enabling effective model compression without sacrificing performance. Extensive experiments on benchmark datasets (Cora, Citeseer, and PubMed) confirm the generalizability of the findings. The implementation is publicly available at: <uri>https://www.github.com/ChuanDai/AdaptedNorm</uri>.https://ieeexplore.ieee.org/document/11059875/GCNlink predictionnode classificationnormalizationRWRTSRT
spellingShingle Chuan Dai
Yajuan Wei
Hao Wang
Ying Liu
Zhijie Xu
AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning Tasks
IEEE Access
GCN
link prediction
node classification
normalization
RWRT
SRT
title AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning Tasks
title_full AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning Tasks
title_fullStr AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning Tasks
title_full_unstemmed AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning Tasks
title_short AdaptedNorm: An Adaptive Modeling Strategy for Graph Convolutional Network-Based Deep Learning Tasks
title_sort adaptednorm an adaptive modeling strategy for graph convolutional network based deep learning tasks
topic GCN
link prediction
node classification
normalization
RWRT
SRT
url https://ieeexplore.ieee.org/document/11059875/
work_keys_str_mv AT chuandai adaptednormanadaptivemodelingstrategyforgraphconvolutionalnetworkbaseddeeplearningtasks
AT yajuanwei adaptednormanadaptivemodelingstrategyforgraphconvolutionalnetworkbaseddeeplearningtasks
AT haowang adaptednormanadaptivemodelingstrategyforgraphconvolutionalnetworkbaseddeeplearningtasks
AT yingliu adaptednormanadaptivemodelingstrategyforgraphconvolutionalnetworkbaseddeeplearningtasks
AT zhijiexu adaptednormanadaptivemodelingstrategyforgraphconvolutionalnetworkbaseddeeplearningtasks