Achieving Adaptive Privacy-Preserving Graph Neural Networks Training in Cloud Environment

Yanli Yuan, Dian Lei, Qing Fan, Keli Zhao, Liehuang Zhu, Chuan Zhang*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

With the widespread adoption of Graph Neural Network (GNN) technology in industry, concerns regarding graph data privacy have become increasingly prominent. Differential privacy has been demonstrated as an effective method to ensure privacy in graph learning. However, existing differential privacy-based GNN methods often overlook the individual privacy protection needs of users, offering uniform privacy guarantees to all. This approach can result in either over-protection or insufficient protection for certain users. To address this issue, we propose an adaptive privacy-preserving GNN training method that accommodates the varying privacy requirements of nodes while achieving high model training accuracy. Specifically, APPGNN allocates adaptive privacy budgets based on individual user privacy needs. Additionally, to mitigate the impact of noise on data utility, APPGNN incorporates a weighted neighborhood aggregation mechanism to enhance GNN model accuracy. Theoretical analysis indicates that APPGNN provides adaptive privacy protection while ensuring ϵ-differential privacy on node data. Experimental evaluations on four real-world graph datasets validate the effectiveness of APPGNN.

Original languageEnglish
Title of host publication2024 IEEE 12th International Conference on Information and Communication Networks, ICICN 2024
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages181-186
Number of pages6
ISBN (Electronic)9798350355802
DOIs
Publication statusPublished - 2024
Event12th IEEE International Conference on Information and Communication Networks, ICICN 2024 - Guilin, China
Duration: 21 Aug 202424 Aug 2024

Publication series

Name2024 IEEE 12th International Conference on Information and Communication Networks, ICICN 2024

Conference

Conference12th IEEE International Conference on Information and Communication Networks, ICICN 2024
Country/TerritoryChina
CityGuilin
Period21/08/2424/08/24

Keywords

  • adaptive
  • cloud computing
  • differential privacy
  • Graph neural networks
  • privacy-preserving

Fingerprint

Dive into the research topics of 'Achieving Adaptive Privacy-Preserving Graph Neural Networks Training in Cloud Environment'. Together they form a unique fingerprint.

Cite this

Yuan, Y., Lei, D., Fan, Q., Zhao, K., Zhu, L., & Zhang, C. (2024). Achieving Adaptive Privacy-Preserving Graph Neural Networks Training in Cloud Environment. In 2024 IEEE 12th International Conference on Information and Communication Networks, ICICN 2024 (pp. 181-186). (2024 IEEE 12th International Conference on Information and Communication Networks, ICICN 2024). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICICN62625.2024.10761771