中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 计算技术研究所
  3. 中国科学院计算技术研究所
  4. 中国科学院计算技术研究所期刊论文
  5. 英文
Revisiting Edge Perturbation for Graph Neural Network in Graph Data Augmentation and Attack

文献类型：期刊论文

作者Liu, Xin2; Zhang, Yuxiang2; Wu, Meng2; Yan, Mingyu2; He, Kun3; Yan, Wei1,2; Pan, Shirui4; Ye, Xiaochun2; Fan, Dongrui2
刊名IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
出版日期2025-07-01
卷号37期号:7页码:4225-4238
关键词Perturbation methods Graph neural networks Image edge detection Data augmentation Accuracy Training Measurement Optimization Detectors World Wide Web Edge perturbation graph neural network graph data augmentation graph data attack
ISSN号1041-4347
DOI10.1109/TKDE.2025.3565306
英文摘要Edge perturbation is a basic method to modify graph structures. It can be categorized into two veins based on their effects on the performance of graph neural networks (GNNs), i.e., graph data augmentation and attack. Surprisingly, both veins of edge perturbation methods employ the same operations, yet yield opposite effects on GNNs' accuracy. A distinct boundary between these methods in using edge perturbation has never been clearly defined. Consequently, inappropriate perturbations may lead to undesirable outcomes, necessitating precise adjustments to achieve desired effects. Therefore, questions of "why edge perturbation has a two-faced effect?" and "what makes edge perturbation flexible and effective?" still remain unanswered. In this paper, we will answer these questions by proposing a unified formulation and establishing a quantizable boundary between two categories of edge perturbation methods. Specifically, we conduct experiments to elucidate the differences and similarities between these methods and theoretically unify the workflow of these methods by casting it to one optimization problem. Then, we devise Edge Priority Detector (EPD) to generate a novel priority metric, bridging these methods up in the workflow. Experiments show that EPD can make augmentation or attack flexibly and achieve comparable or superior performance to other counterparts with less time overhead.
资助项目National Key Research and Development Program[2023YFB4502305] ; National Natural Science Foundation of China[62202451] ; National Natural Science Foundation of China[62302477] ; CAS Project for Young Scientists in Basic Research[YSBR-029] ; CAS Project for Youth Innovation Promotion Association
WOS研究方向Computer Science ; Engineering
语种英语
WOS记录号WOS:001504151700027
出版者IEEE COMPUTER SOC
源URL[http://119.78.100.204/handle/2XEOYT63/42351]  
专题中国科学院计算技术研究所期刊论文_英文
通讯作者Yan, Mingyu
作者单位1.Zhongguancun Lab, Beijing 100086, Peoples R China
2.Chinese Acad Sci, Inst Comp Technol, SKLP, Beijing 100086, Peoples R China
3.Renmin Univ China, Beijing 100872, Peoples R China
4.Griffith Univ, Brisbane, Qld 4111, Australia
推荐引用方式
GB/T 7714		 Liu, Xin,Zhang, Yuxiang,Wu, Meng,et al. Revisiting Edge Perturbation for Graph Neural Network in Graph Data Augmentation and Attack[J]. IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING,2025,37(7):4225-4238.
APA Liu, Xin.,Zhang, Yuxiang.,Wu, Meng.,Yan, Mingyu.,He, Kun.,...&Fan, Dongrui.(2025).Revisiting Edge Perturbation for Graph Neural Network in Graph Data Augmentation and Attack.IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING,37(7),4225-4238.
MLA Liu, Xin,et al."Revisiting Edge Perturbation for Graph Neural Network in Graph Data Augmentation and Attack".IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 37.7(2025):4225-4238.

入库方式： OAI收割

来源：计算技术研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。