arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.12250cs.LGcs.CRcs.SI
收藏

基于无监督图学习聚类的社区隐蔽

Community Concealment from Graph Neural Networks

Indiana University(印第安纳大学);Oak Ridge National Laboratory(橡树岭国家实验室);UNC Charlotte(北卡罗来纳大学夏洛特分校)

Dalyapraz Manatova, Pablo Moriano, L. Jean Camp

AI总结:

本文提出了一种基于图神经网络的社区隐蔽方法,通过扰动策略减少社区区分性,提升隐私保护效果。

摘要:

图神经网络(GNN)被设计用于利用带属性的图来学习表示。这些表示在无监督学习的聚类和社区检测中是有益的。然而,这种推断可能会揭示敏感群体、聚类系统或集体行为,从而引发关于群体层面隐私的担忧。例如,在社会和关键基础设施网络中的社区属性,可能会暴露协调资产组、运营层次结构和系统依赖性,这些可能被用于画像或情报收集。我们研究了一种防御场景,其中数据发布者(防御者)试图隐藏感兴趣的社区,同时在网络中做出有限的、具有实用性的修改。我们的分析表明,社区隐蔽性受到两个可量化的因素强烈影响:社区边界处的连通性以及受保护社区与相邻社区之间的特征相似性。基于这些发现,我们提出了一种扰动策略,通过重 wiring 选定的边并修改节点特征,以减少 GNN 消息传递中利用的区分性。在合成基准和真实网络图上,所提出的方法在相同的扰动预算下优于 DICE。总体而言,在评估的设置中,它实现了约 20-45% 的中位相对隐蔽性改进。这些发现展示了一种对抗 GNN 基础的社区学习的缓解策略,并突显了图学习中固有的群体层面隐私风险。

英文摘要:

Graph neural networks (GNNs) enable powerful unsupervised learning of communities. However, such inference may inadvertently expose sensitive group structures, critical clustered patterns, or collective behaviors, raising concerns about sensitive group-level privacy. In social and critical infrastructure networks, unauthorized community inference can reveal coordinated asset groups, operational hierarchies, and system dependencies that may be exploited for reconnaissance or profiling. We study a defensive setting in which a network (or defender) operator seeks to conceal a community of interest while making only small, utility-preserving modifications to the network. Our analysis shows that community concealment depends on two measurable factors: the connectivity at the community boundary and the feature similarity between the protected community and its neighbors. Guided by these observations, we introduce Feature-Community-guided DICE (FCom-DICE), a perturbation strategy built on DICE (Disconnect Internally Connect Externally) that rewires a set of structurally influential edges and adjusts node features to reduce the distinctiveness exploited by GNN message passing. Across synthetic benchmarks and real network graphs such as Facebook, Wikipedia, and Bitcoin Transactions, FCom-DICE consistently outperforms structure-only DICE under the same perturbation budgets. The largest improvements are observed for communities that are weakly connected to the rest of the network and well separated in feature space. These gains are achieved while preserving key structural and feature characteristics of the original network. These results demonstrate the effectiveness of feature-aware perturbations for reducing the recoverability of targeted communities under GNN-based community inference.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑