Saved in:
Bibliographic Details
Main Author: Pandey, Daksh
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.25205
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918150863323136
author Pandey, Daksh
author_facet Pandey, Daksh
contents Self-supervised learning (SSL) has emerged as a powerful paradigm for learning representations on graph data without requiring manual labels. However, leading SSL methods like GRACE are fundamentally incompatible with privacy-preserving technologies such as Homomorphic Encryption (HE) due to their reliance on non-polynomial operations. This paper introduces Poly-GRACE, a novel framework for HE-compatible self-supervised learning on graphs. Our approach consists of a fully polynomial-friendly Graph Convolutional Network (GCN) encoder and a novel, polynomial-based contrastive loss function. Through experiments on three benchmark datasets -- Cora, CiteSeer, and PubMed -- we demonstrate that Poly-GRACE not only enables private pre-training but also achieves performance that is highly competitive with, and in the case of CiteSeer, superior to the standard non-private baseline. Our work represents a significant step towards practical and high-performance privacy-preserving graph representation learning.
format Preprint
id arxiv_https___arxiv_org_abs_2509_25205
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Polynomial Contrastive Learning for Privacy-Preserving Representation Learning on Graphs
Pandey, Daksh
Machine Learning
Cryptography and Security
Rings and Algebras
Self-supervised learning (SSL) has emerged as a powerful paradigm for learning representations on graph data without requiring manual labels. However, leading SSL methods like GRACE are fundamentally incompatible with privacy-preserving technologies such as Homomorphic Encryption (HE) due to their reliance on non-polynomial operations. This paper introduces Poly-GRACE, a novel framework for HE-compatible self-supervised learning on graphs. Our approach consists of a fully polynomial-friendly Graph Convolutional Network (GCN) encoder and a novel, polynomial-based contrastive loss function. Through experiments on three benchmark datasets -- Cora, CiteSeer, and PubMed -- we demonstrate that Poly-GRACE not only enables private pre-training but also achieves performance that is highly competitive with, and in the case of CiteSeer, superior to the standard non-private baseline. Our work represents a significant step towards practical and high-performance privacy-preserving graph representation learning.
title Polynomial Contrastive Learning for Privacy-Preserving Representation Learning on Graphs
topic Machine Learning
Cryptography and Security
Rings and Algebras
url https://arxiv.org/abs/2509.25205