Locally Differentially Private Graph Clustering via the Power Iteration Method

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Suppakitpaisarn, Vorapong, Mukherjee, Sayan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918022212485120
author Suppakitpaisarn, Vorapong
Mukherjee, Sayan
author_facet Suppakitpaisarn, Vorapong
Mukherjee, Sayan
contents We propose a locally differentially private graph clustering algorithm. Previous works have explored this problem, including approaches that apply spectral clustering to graphs generated via the randomized response algorithm. However, these methods only achieve accurate results when the privacy budget is in $Ω(\log n)$, which is unsuitable for many practical applications. In response, we present an interactive algorithm based on the power iteration method. Given that the noise introduced by the largest eigenvector constant can be significant, we incorporate a technique to eliminate this constant. As a result, our algorithm attains local differential privacy with a constant privacy budget when the graph is well-clustered and has a minimum degree of $\tildeΩ(\sqrt{n})$. In contrast, while randomized response has been shown to produce accurate results under the same minimum degree condition, it is limited to graphs generated from the stochastic block model. We perform experiments to demonstrate that our method outperforms spectral clustering applied to randomized response results.
format Preprint
id arxiv_https___arxiv_org_abs_2505_11169
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Locally Differentially Private Graph Clustering via the Power Iteration Method
Suppakitpaisarn, Vorapong
Mukherjee, Sayan
Data Structures and Algorithms
Social and Information Networks
We propose a locally differentially private graph clustering algorithm. Previous works have explored this problem, including approaches that apply spectral clustering to graphs generated via the randomized response algorithm. However, these methods only achieve accurate results when the privacy budget is in $Ω(\log n)$, which is unsuitable for many practical applications. In response, we present an interactive algorithm based on the power iteration method. Given that the noise introduced by the largest eigenvector constant can be significant, we incorporate a technique to eliminate this constant. As a result, our algorithm attains local differential privacy with a constant privacy budget when the graph is well-clustered and has a minimum degree of $\tildeΩ(\sqrt{n})$. In contrast, while randomized response has been shown to produce accurate results under the same minimum degree condition, it is limited to graphs generated from the stochastic block model. We perform experiments to demonstrate that our method outperforms spectral clustering applied to randomized response results.
title Locally Differentially Private Graph Clustering via the Power Iteration Method
topic Data Structures and Algorithms
Social and Information Networks
url https://arxiv.org/abs/2505.11169