GraphCloak: Safeguarding Task-specific Knowledge within Graph-structured Data from Unauthorized Exploitation

As Graph Neural Networks (GNNs) become increasingly prevalent in a variety of fields, from social network analysis to protein-protein interaction studies, growing concerns have emerged regarding the unauthorized utilization of personal data. Recent studies have shown that imperceptible poisoning att...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2023-10
Hauptverfasser: Liu, Yixin, Fan, Chenrui, Chen, Xun, Zhou, Pan, Sun, Lichao
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Liu, Yixin
Fan, Chenrui
Chen, Xun
Zhou, Pan
Sun, Lichao
description As Graph Neural Networks (GNNs) become increasingly prevalent in a variety of fields, from social network analysis to protein-protein interaction studies, growing concerns have emerged regarding the unauthorized utilization of personal data. Recent studies have shown that imperceptible poisoning attacks are an effective method of protecting image data from such misuse. However, the efficacy of this approach in the graph domain remains unexplored. To bridge this gap, this paper introduces GraphCloak to safeguard against the unauthorized usage of graph data. Compared with prior work, GraphCloak offers unique significant innovations: (1) graph-oriented, the perturbations are applied to both topological structures and descriptive features of the graph; (2) effective and stealthy, our cloaking method can bypass various inspections while causing a significant performance drop in GNNs trained on the cloaked graphs; and (3) stable across settings, our methods consistently perform effectively under a range of practical settings with limited knowledge. To address the intractable bi-level optimization problem, we propose two error-minimizing-based poisoning methods that target perturbations on the structural and feature space, along with a subgraph injection poisoning method. Our comprehensive evaluation of these methods underscores their effectiveness, stealthiness, and stability. We also delve into potential countermeasures and provide analytical justification for their effectiveness, paving the way for intriguing future research.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2876196980</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2876196980</sourcerecordid><originalsourceid>FETCH-proquest_journals_28761969803</originalsourceid><addsrcrecordid>eNqNi8sKgkAYRocgSMp3GGgt6Ji3tmYFLau1_Oioo9OMzQWjp0-iB2j1wTnnWyCHhGHgpTtCVsjVuvd9n8QJiaLQQfykYOxyLmHY4ys0tLWgaiZafAM9eHqkFWtYhS9CTpzWLcUTMx0T-PvztFG2MlbRGh_AAG6UfOC7AGs6qdh7xsVr5JIZMEyKDVo2wDV1f7tG22Nxy8_eqOTTUm3KXlolZlWSNImDLM5SP_yv-gDvqEpE</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2876196980</pqid></control><display><type>article</type><title>GraphCloak: Safeguarding Task-specific Knowledge within Graph-structured Data from Unauthorized Exploitation</title><source>Free E- Journals</source><creator>Liu, Yixin ; Fan, Chenrui ; Chen, Xun ; Zhou, Pan ; Sun, Lichao</creator><creatorcontrib>Liu, Yixin ; Fan, Chenrui ; Chen, Xun ; Zhou, Pan ; Sun, Lichao</creatorcontrib><description>As Graph Neural Networks (GNNs) become increasingly prevalent in a variety of fields, from social network analysis to protein-protein interaction studies, growing concerns have emerged regarding the unauthorized utilization of personal data. Recent studies have shown that imperceptible poisoning attacks are an effective method of protecting image data from such misuse. However, the efficacy of this approach in the graph domain remains unexplored. To bridge this gap, this paper introduces GraphCloak to safeguard against the unauthorized usage of graph data. Compared with prior work, GraphCloak offers unique significant innovations: (1) graph-oriented, the perturbations are applied to both topological structures and descriptive features of the graph; (2) effective and stealthy, our cloaking method can bypass various inspections while causing a significant performance drop in GNNs trained on the cloaked graphs; and (3) stable across settings, our methods consistently perform effectively under a range of practical settings with limited knowledge. To address the intractable bi-level optimization problem, we propose two error-minimizing-based poisoning methods that target perturbations on the structural and feature space, along with a subgraph injection poisoning method. Our comprehensive evaluation of these methods underscores their effectiveness, stealthiness, and stability. We also delve into potential countermeasures and provide analytical justification for their effectiveness, paving the way for intriguing future research.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Effectiveness ; Graph neural networks ; Graph theory ; Network analysis ; Optimization ; Perturbation ; Poisoning ; Proteins ; Social networks ; Stability analysis ; Structured data</subject><ispartof>arXiv.org, 2023-10</ispartof><rights>2023. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>776,780</link.rule.ids></links><search><creatorcontrib>Liu, Yixin</creatorcontrib><creatorcontrib>Fan, Chenrui</creatorcontrib><creatorcontrib>Chen, Xun</creatorcontrib><creatorcontrib>Zhou, Pan</creatorcontrib><creatorcontrib>Sun, Lichao</creatorcontrib><title>GraphCloak: Safeguarding Task-specific Knowledge within Graph-structured Data from Unauthorized Exploitation</title><title>arXiv.org</title><description>As Graph Neural Networks (GNNs) become increasingly prevalent in a variety of fields, from social network analysis to protein-protein interaction studies, growing concerns have emerged regarding the unauthorized utilization of personal data. Recent studies have shown that imperceptible poisoning attacks are an effective method of protecting image data from such misuse. However, the efficacy of this approach in the graph domain remains unexplored. To bridge this gap, this paper introduces GraphCloak to safeguard against the unauthorized usage of graph data. Compared with prior work, GraphCloak offers unique significant innovations: (1) graph-oriented, the perturbations are applied to both topological structures and descriptive features of the graph; (2) effective and stealthy, our cloaking method can bypass various inspections while causing a significant performance drop in GNNs trained on the cloaked graphs; and (3) stable across settings, our methods consistently perform effectively under a range of practical settings with limited knowledge. To address the intractable bi-level optimization problem, we propose two error-minimizing-based poisoning methods that target perturbations on the structural and feature space, along with a subgraph injection poisoning method. Our comprehensive evaluation of these methods underscores their effectiveness, stealthiness, and stability. We also delve into potential countermeasures and provide analytical justification for their effectiveness, paving the way for intriguing future research.</description><subject>Effectiveness</subject><subject>Graph neural networks</subject><subject>Graph theory</subject><subject>Network analysis</subject><subject>Optimization</subject><subject>Perturbation</subject><subject>Poisoning</subject><subject>Proteins</subject><subject>Social networks</subject><subject>Stability analysis</subject><subject>Structured data</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>BENPR</sourceid><recordid>eNqNi8sKgkAYRocgSMp3GGgt6Ji3tmYFLau1_Oioo9OMzQWjp0-iB2j1wTnnWyCHhGHgpTtCVsjVuvd9n8QJiaLQQfykYOxyLmHY4ys0tLWgaiZafAM9eHqkFWtYhS9CTpzWLcUTMx0T-PvztFG2MlbRGh_AAG6UfOC7AGs6qdh7xsVr5JIZMEyKDVo2wDV1f7tG22Nxy8_eqOTTUm3KXlolZlWSNImDLM5SP_yv-gDvqEpE</recordid><startdate>20231011</startdate><enddate>20231011</enddate><creator>Liu, Yixin</creator><creator>Fan, Chenrui</creator><creator>Chen, Xun</creator><creator>Zhou, Pan</creator><creator>Sun, Lichao</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20231011</creationdate><title>GraphCloak: Safeguarding Task-specific Knowledge within Graph-structured Data from Unauthorized Exploitation</title><author>Liu, Yixin ; Fan, Chenrui ; Chen, Xun ; Zhou, Pan ; Sun, Lichao</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_28761969803</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Effectiveness</topic><topic>Graph neural networks</topic><topic>Graph theory</topic><topic>Network analysis</topic><topic>Optimization</topic><topic>Perturbation</topic><topic>Poisoning</topic><topic>Proteins</topic><topic>Social networks</topic><topic>Stability analysis</topic><topic>Structured data</topic><toplevel>online_resources</toplevel><creatorcontrib>Liu, Yixin</creatorcontrib><creatorcontrib>Fan, Chenrui</creatorcontrib><creatorcontrib>Chen, Xun</creatorcontrib><creatorcontrib>Zhou, Pan</creatorcontrib><creatorcontrib>Sun, Lichao</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Liu, Yixin</au><au>Fan, Chenrui</au><au>Chen, Xun</au><au>Zhou, Pan</au><au>Sun, Lichao</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>GraphCloak: Safeguarding Task-specific Knowledge within Graph-structured Data from Unauthorized Exploitation</atitle><jtitle>arXiv.org</jtitle><date>2023-10-11</date><risdate>2023</risdate><eissn>2331-8422</eissn><abstract>As Graph Neural Networks (GNNs) become increasingly prevalent in a variety of fields, from social network analysis to protein-protein interaction studies, growing concerns have emerged regarding the unauthorized utilization of personal data. Recent studies have shown that imperceptible poisoning attacks are an effective method of protecting image data from such misuse. However, the efficacy of this approach in the graph domain remains unexplored. To bridge this gap, this paper introduces GraphCloak to safeguard against the unauthorized usage of graph data. Compared with prior work, GraphCloak offers unique significant innovations: (1) graph-oriented, the perturbations are applied to both topological structures and descriptive features of the graph; (2) effective and stealthy, our cloaking method can bypass various inspections while causing a significant performance drop in GNNs trained on the cloaked graphs; and (3) stable across settings, our methods consistently perform effectively under a range of practical settings with limited knowledge. To address the intractable bi-level optimization problem, we propose two error-minimizing-based poisoning methods that target perturbations on the structural and feature space, along with a subgraph injection poisoning method. Our comprehensive evaluation of these methods underscores their effectiveness, stealthiness, and stability. We also delve into potential countermeasures and provide analytical justification for their effectiveness, paving the way for intriguing future research.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2023-10
issn 2331-8422
language eng
recordid cdi_proquest_journals_2876196980
source Free E- Journals
subjects Effectiveness
Graph neural networks
Graph theory
Network analysis
Optimization
Perturbation
Poisoning
Proteins
Social networks
Stability analysis
Structured data
title GraphCloak: Safeguarding Task-specific Knowledge within Graph-structured Data from Unauthorized Exploitation
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-01T14%3A40%3A32IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=GraphCloak:%20Safeguarding%20Task-specific%20Knowledge%20within%20Graph-structured%20Data%20from%20Unauthorized%20Exploitation&rft.jtitle=arXiv.org&rft.au=Liu,%20Yixin&rft.date=2023-10-11&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2876196980%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2876196980&rft_id=info:pmid/&rfr_iscdi=true