Dynamic Node Privacy Feature Decoupling Graph Autoencoder Based on Attention Mechanism

Graph autoencoders’ inherent capability to capture node feature correlations poses significant privacy risks through attackers inference. Previous feature decoupling approaches predominantly apply uniform privacy protection across nodes, disregarding the varying sensitivity levels inherent in graph...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Applied sciences Ročník 15; číslo 12; s. 6489
Hlavní autoři: Huang, Yikai, Tang, Jinchuan, Dang, Shuping
Médium: Journal Article
Jazyk:angličtina
Vydáno: Basel MDPI AG 01.06.2025
Témata:
ISSN:2076-3417, 2076-3417
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Graph autoencoders’ inherent capability to capture node feature correlations poses significant privacy risks through attackers inference. Previous feature decoupling approaches predominantly apply uniform privacy protection across nodes, disregarding the varying sensitivity levels inherent in graph structures. To solve the above problems, we propose a novel dual-path graph autoencoder incorporating attention-aware privacy adaptation. Firstly, we design an attention-driven metric learning framework to quantify node-specific privacy importance through attention weights and select important nodes to construct the privacy distribution, so that realizing the dynamically privacy decoupling and reducing utility loss. Then, we introduce Hilbert-Schmidt Independence Criterion (HSIC) to measure the dependence between privacy and non-privacy information, which avoids the deviations that occur when using approximate methods such as variational inference. Finally, we use the method of alternating training to comprehensively evaluate the privacy importance of nodes. Experimental results on three real-world datasets—Yale, Rochester, and Credit Defaulter—demonstrate that our proposed method significantly outperforms existing approaches like PVGAE, GAE-MI, and APGE, where the inference accuracy regarding privacy decreased by 25.5%, but the accuracy rate of link prediction achieved the highest 84.7% compared to other methods.
AbstractList Graph autoencoders’ inherent capability to capture node feature correlations poses significant privacy risks through attackers inference. Previous feature decoupling approaches predominantly apply uniform privacy protection across nodes, disregarding the varying sensitivity levels inherent in graph structures. To solve the above problems, we propose a novel dual-path graph autoencoder incorporating attention-aware privacy adaptation. Firstly, we design an attention-driven metric learning framework to quantify node-specific privacy importance through attention weights and select important nodes to construct the privacy distribution, so that realizing the dynamically privacy decoupling and reducing utility loss. Then, we introduce Hilbert-Schmidt Independence Criterion (HSIC) to measure the dependence between privacy and non-privacy information, which avoids the deviations that occur when using approximate methods such as variational inference. Finally, we use the method of alternating training to comprehensively evaluate the privacy importance of nodes. Experimental results on three real-world datasets—Yale, Rochester, and Credit Defaulter—demonstrate that our proposed method significantly outperforms existing approaches like PVGAE, GAE-MI, and APGE, where the inference accuracy regarding privacy decreased by 25.5%, but the accuracy rate of link prediction achieved the highest 84.7% compared to other methods.
Audience Academic
Author Tang, Jinchuan
Dang, Shuping
Huang, Yikai
Author_xml – sequence: 1
  givenname: Yikai
  surname: Huang
  fullname: Huang, Yikai
– sequence: 2
  givenname: Jinchuan
  surname: Tang
  fullname: Tang, Jinchuan
– sequence: 3
  givenname: Shuping
  surname: Dang
  fullname: Dang, Shuping
BookMark eNpNUctuFDEQtFCQCCEnfsASR7TBzxn7uCQkRAokB-Bq9dg9G6927cHjQdq_j8kilO5Dt0rVpVLXW3KSckJC3nN2IaVln2CauOaiU8a-IqeC9d1KKt6fvNjfkPN53rJWlkvD2Sn5dXVIsI-efs8B6UOJf8Af6DVCXQrSK_R5mXYxbehNgemRrpeaMfnGLfQzzBhoTnRdK6Ya2_YN_SOkOO_fkdcj7GY8_zfPyM_rLz8uv67u7m9uL9d3Ky87WVe91Ur1oFCwIQzDqEEzEZQFAyAM91xC50evlO1UkHqUtu9MY_c2GDaMgzwjt0fdkGHrphL3UA4uQ3TPQC4bB6VGv0NnlQRQelSqvUgFaxQwwCC4DygkYtP6cNSaSv694FzdNi8lNftOCiFNL4WWjXVxZG2gicY05lrAtw7Y3tgSGWPD10Zp3TNjdDv4eDzwJc9zwfG_Tc7c3-Dci-DkE1kJi4I
Cites_doi 10.1142/S0218488502001648
10.1016/j.neucom.2024.129001
10.1007/11787006_1
10.1145/3534678.3539321
10.1145/1055558.1055591
10.1145/3534678.3539302
10.1145/3583780.3614933
10.1007/s00355-023-01456-4
10.1016/j.knosys.2021.107567
10.1109/JIOT.2020.3036583
10.1109/TDSC.2024.3417513
10.1145/3447548.3467273
10.3390/su151511893
10.1145/3637528.3672013
10.1145/3437963.3441752
10.1145/3448891.3448939
10.1145/3534678.3539232
10.1007/s10207-022-00646-y
10.1109/IJCNN55064.2022.9892789
ContentType Journal Article
Copyright COPYRIGHT 2025 MDPI AG
2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Copyright_xml – notice: COPYRIGHT 2025 MDPI AG
– notice: 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
DBID AAYXX
CITATION
ABUWG
AFKRA
AZQEC
BENPR
CCPQU
DWQXO
PHGZM
PHGZT
PIMPY
PKEHL
PQEST
PQQKQ
PQUKI
PRINS
DOA
DOI 10.3390/app15126489
DatabaseName CrossRef
ProQuest Central (Alumni)
ProQuest Central UK/Ireland
ProQuest Central Essentials - QC
ProQuest Central
ProQuest One
ProQuest Central
ProQuest Central Premium
ProQuest One Academic
Publicly Available Content Database (ProQuest)
ProQuest One Academic Middle East (New)
ProQuest One Academic Eastern Edition (DO NOT USE)
ProQuest One Academic (retired)
ProQuest One Academic UKI Edition
ProQuest Central China
DOAJ Directory of Open Access Journals
DatabaseTitle CrossRef
Publicly Available Content Database
ProQuest One Academic Middle East (New)
ProQuest Central Essentials
ProQuest One Academic Eastern Edition
ProQuest Central (Alumni Edition)
ProQuest One Community College
ProQuest Central China
ProQuest Central
ProQuest One Academic UKI Edition
ProQuest Central Korea
ProQuest Central (New)
ProQuest One Academic
ProQuest One Academic (New)
DatabaseTitleList Publicly Available Content Database

CrossRef

Database_xml – sequence: 1
  dbid: DOA
  name: DOAJ Directory of Open Access Journals
  url: https://www.doaj.org/
  sourceTypes: Open Website
– sequence: 2
  dbid: PIMPY
  name: Publicly Available Content Database
  url: http://search.proquest.com/publiccontent
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
Sciences (General)
EISSN 2076-3417
ExternalDocumentID oai_doaj_org_article_943aa45f442644d984a0aed21cde23ee
A845570885
10_3390_app15126489
GroupedDBID .4S
2XV
5VS
7XC
8CJ
8FE
8FG
8FH
AADQD
AAFWJ
AAYXX
ADBBV
ADMLS
AFFHD
AFKRA
AFPKN
AFZYC
ALMA_UNASSIGNED_HOLDINGS
APEBS
ARCSS
BCNDV
BENPR
CCPQU
CITATION
CZ9
D1I
D1J
D1K
GROUPED_DOAJ
IAO
IGS
ITC
K6-
K6V
KC.
KQ8
L6V
LK5
LK8
M7R
MODMG
M~E
OK1
P62
PHGZM
PHGZT
PIMPY
PROAC
TUS
ABUWG
AZQEC
DWQXO
PKEHL
PQEST
PQQKQ
PQUKI
PRINS
ID FETCH-LOGICAL-c363t-795447a4e20bdbbf5a502d49a8aa281c13a6cfc44964d35f397684e279d80bfb3
IEDL.DBID PIMPY
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001515296400001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2076-3417
IngestDate Tue Oct 14 18:45:32 EDT 2025
Mon Jun 30 07:18:31 EDT 2025
Tue Nov 04 18:14:56 EST 2025
Sat Nov 29 07:12:53 EST 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 12
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c363t-795447a4e20bdbbf5a502d49a8aa281c13a6cfc44964d35f397684e279d80bfb3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
OpenAccessLink https://www.proquest.com/publiccontent/docview/3223873253?pq-origsite=%requestingapplication%
PQID 3223873253
PQPubID 2032433
ParticipantIDs doaj_primary_oai_doaj_org_article_943aa45f442644d984a0aed21cde23ee
proquest_journals_3223873253
gale_infotracacademiconefile_A845570885
crossref_primary_10_3390_app15126489
PublicationCentury 2000
PublicationDate 2025-06-01
PublicationDateYYYYMMDD 2025-06-01
PublicationDate_xml – month: 06
  year: 2025
  text: 2025-06-01
  day: 01
PublicationDecade 2020
PublicationPlace Basel
PublicationPlace_xml – name: Basel
PublicationTitle Applied sciences
PublicationYear 2025
Publisher MDPI AG
Publisher_xml – name: MDPI AG
References Sweeney (ref_8) 2002; 10
Wang (ref_5) 2023; 22
ref_14
ref_13
Lin (ref_17) 2024; 22
ref_12
ref_11
ref_10
Amara (ref_2) 2025; 619
ref_19
ref_18
ref_16
ref_15
Wang (ref_7) 2021; 234
Bloch (ref_23) 2023; 61
ref_24
ref_22
ref_21
ref_20
ref_1
ref_3
Li (ref_25) 2020; 8
ref_9
ref_4
ref_6
References_xml – volume: 10
  start-page: 557
  year: 2002
  ident: ref_8
  article-title: k-anonymity: A model for protecting privacy
  publication-title: Int. J. Uncertain. Fuzziness- Knowl.-Based Syst.
  doi: 10.1142/S0218488502001648
– volume: 619
  start-page: 129001
  year: 2025
  ident: ref_2
  article-title: A multi-view GNN-based network representation learning framework for recommendation systems
  publication-title: Neurocomputing
  doi: 10.1016/j.neucom.2024.129001
– ident: ref_10
  doi: 10.1007/11787006_1
– ident: ref_11
– ident: ref_1
– ident: ref_4
  doi: 10.1145/3534678.3539321
– ident: ref_9
  doi: 10.1145/1055558.1055591
– ident: ref_15
  doi: 10.1145/3534678.3539302
– ident: ref_20
  doi: 10.1145/3583780.3614933
– ident: ref_21
– volume: 61
  start-page: 413
  year: 2023
  ident: ref_23
  article-title: Centrality measures in networks
  publication-title: Soc. Choice Welf.
  doi: 10.1007/s00355-023-01456-4
– ident: ref_6
– volume: 234
  start-page: 107567
  year: 2021
  ident: ref_7
  article-title: Learning with Hilbert–Schmidt independence criterion: A review and new perspectives
  publication-title: Knowl.-Based Syst.
  doi: 10.1016/j.knosys.2021.107567
– volume: 8
  start-page: 6904
  year: 2020
  ident: ref_25
  article-title: Adversarial privacy-preserving graph embedding against inference attack
  publication-title: IEEE Internet Things J.
  doi: 10.1109/JIOT.2020.3036583
– volume: 22
  start-page: 967
  year: 2024
  ident: ref_17
  article-title: Graph privacy funnel: A variational approach for privacy-preserving representation learning on graphs
  publication-title: IEEE Trans. Dependable Secur. Comput.
  doi: 10.1109/TDSC.2024.3417513
– ident: ref_24
  doi: 10.1145/3447548.3467273
– ident: ref_3
  doi: 10.3390/su151511893
– ident: ref_13
– ident: ref_18
  doi: 10.1145/3637528.3672013
– ident: ref_14
  doi: 10.1145/3437963.3441752
– ident: ref_19
  doi: 10.1145/3448891.3448939
– ident: ref_16
  doi: 10.1145/3534678.3539232
– ident: ref_22
– volume: 22
  start-page: 497
  year: 2023
  ident: ref_5
  article-title: Defense against membership inference attack in graph neural networks through graph perturbation
  publication-title: Int. J. Inf. Secur.
  doi: 10.1007/s10207-022-00646-y
– ident: ref_12
  doi: 10.1109/IJCNN55064.2022.9892789
SSID ssj0000913810
Score 2.320442
Snippet Graph autoencoders’ inherent capability to capture node feature correlations poses significant privacy risks through attackers inference. Previous feature...
SourceID doaj
proquest
gale
crossref
SourceType Open Website
Aggregation Database
Index Database
StartPage 6489
SubjectTerms attention
graph autoencoder
Hypothesis testing
Methods
Neural networks
Privacy
privacy decouple
Privacy, Right of
privacy-utility trade-off
Social networks
SummonAdditionalLinks – databaseName: DOAJ Directory of Open Access Journals
  dbid: DOA
  link: http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1Nb9QwEB2higMcEC0gFgryoVLhEJHYk8Q-biltL131AKg3a5Kx0R7YrbLZSvx7PEmKckFcuEY-WG8yX7bnPYATjCnsMdmMcsMZNsyZpSZmzJUJddCVRh7EJurVyt7eupuZ1Je8CRvpgUfgPjk0RFhGHFI3O4uUU2BdtBy0CUGib6p6Zs3UEINdIdRV40CeSX293AdLcqtQBN1nKWhg6v9bPB6SzMVzeDZVh2o57uoQHoXNETydcQYeweHkjTv1YaKM_vgCvp-PwvJqteWgbrr1PbW_lJR3-y6o89Ri7mXy9oe6FH5qtdz3WyGw5NCps5TGWG03atn349NHdR1kHHi9-_kSvl18-fr5KpsUE7LWVKbPalci1oRB5w03TSypzDWjI0ukbdEWhqo2toiuQjZllGLEptW1Y5s3sTGv4GCz3YTXoBLGcu-bV6QDVqxdKgVjEXOk5OSBzAJOHkD0dyMxhk8NhWDtZ1gv4EwA_rNE2KyHD8nGfrKx_5eNF3Aq5vHic31HLU2jA2mnwl7llxaFSczacgHHDxb0kzPufIpZxtZGl-bN_9jNW3iiRQR4OIo5hoO-24d38Li979e77v3wH_4GOFPicw
  priority: 102
  providerName: Directory of Open Access Journals
Title Dynamic Node Privacy Feature Decoupling Graph Autoencoder Based on Attention Mechanism
URI https://www.proquest.com/docview/3223873253
https://doaj.org/article/943aa45f442644d984a0aed21cde23ee
Volume 15
WOSCitedRecordID wos001515296400001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVAON
  databaseName: DOAJ Directory of Open Access Journals
  customDbUrl:
  eissn: 2076-3417
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000913810
  issn: 2076-3417
  databaseCode: DOA
  dateStart: 20110101
  isFulltext: true
  titleUrlDefault: https://www.doaj.org/
  providerName: Directory of Open Access Journals
– providerCode: PRVHPJ
  databaseName: ROAD: Directory of Open Access Scholarly Resources
  customDbUrl:
  eissn: 2076-3417
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000913810
  issn: 2076-3417
  databaseCode: M~E
  dateStart: 20110101
  isFulltext: true
  titleUrlDefault: https://road.issn.org
  providerName: ISSN International Centre
– providerCode: PRVPQU
  databaseName: ProQuest Central
  customDbUrl:
  eissn: 2076-3417
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000913810
  issn: 2076-3417
  databaseCode: BENPR
  dateStart: 20110101
  isFulltext: true
  titleUrlDefault: https://www.proquest.com/central
  providerName: ProQuest
– providerCode: PRVPQU
  databaseName: Publicly Available Content Database
  customDbUrl:
  eissn: 2076-3417
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000913810
  issn: 2076-3417
  databaseCode: PIMPY
  dateStart: 20110101
  isFulltext: true
  titleUrlDefault: http://search.proquest.com/publiccontent
  providerName: ProQuest
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1Nb9QwEB3BLgc4AC0gFsrKh0rAIWpiO4lzQru0BQ5dRQhQOUWTjFPtgU1JspX493gSb9kLnLgmlmLleT489rwHcKxr5_YITYChokCXRIHBsg6IEmVTKxOpaRCbSFcrc3mZ5b49uvPXKnc-cXDUI9sz39t2TviEmoor5iduGSqTKhmrd9c_A9aQ4rNWL6hxF6ZMvGUmMM0_XeTfb2suzIFponBs01Nut8-nxBzyEs0y73uBaeDv_5uXHkLP-aP_O-nH8NCnoGIxrpkDuGM3h_Bgj5jwEA68yXfijeelfvsEvp2O6vVi1ZAVebu-weqX4Bxy21px6vaxW27vvRIfmARbLLZ9wyyZZFuxdLGSRLMRi74f71eKC8s9x-vux1P4en725f3HwMsyBJVKVB-kmfu5KWorw5LKso4xDiXpDA2iNFEVKUyqutI6SzSpuOaMx7jRaUYmLOtSPYPJptnY5yDQEh8uhwlKqxOSmcs366gONTpPYlHN4HiHSXE9sm8UbtfC0BV70M1gyXjdDmHK7OFB014V3gKLTCtEHdd6yAEpMxpD930ZVWSlsnYGrxntgg27b7FC35_gZsoUWcXCaKYrMyaewdEO7cJbfFf8AffFv1-_hPuSNYSHSs4RTPp2a1_BveqmX3ftHKbLs1X-eT7UBuZ-Af8GPnP_mA
linkProvider ProQuest
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMw1V3LbtQwFL0qUyTKAmgBdaCAF0XAIiKxncRZIDRlKB21M5pFQe3KOLZTzaJJSTJF81N8I755lNnArgu2iZWXT869ftxzAPZ55mjPKOEpnxmPp8Z4QqWZZ0zEbGxpRLlpzCbi2UycnSXzDfjV18LgtsqeExuiNoXGOfL3DnhMxIyG7OPVDw9do3B1tbfQaGFxbFc_3ZCt-jAZu_59Tenh59NPR17nKuBpFrHai5OQ81hxS_3UpGkWqtCnhidKKEVFoAOmIp1pzpOIGxZmGLCFax0nRvhpljJ33TuwyR3YxQA255Pp_PxmVgdVNkXgt4WAjCU-rkNjUI04Gsmvhb7GIeBvcaAJbocP_7fP8ggedGk0GbW434YNm-_A_TVxxR3Y7mirIm87be13j-HbeJWry4Ums8JYMi8X10qvCObBy9KSsRuLL7FE-YJ8QSFvMlrWBSp9GluSAxfvDSlyMqrrdo8omVqsm15Ul0_g66287VMY5EVud4Eoa3CB3I8UtTwyNHE5cxZkPleODa1iQ9jve11etQoi0o28EBxyDRxDOEBE3DRB2e_mQFFeyI5FZMKZUjzMeJPHmkRw5bv700AbS5m1Q3iDeJJITnWptOpqLNyTosyXHAmOkmtChEPY6_EkO9aq5B8wPfv36Vdw7-h0eiJPJrPj57BF0RO5mZnag0FdLu0LuKuv60VVvux-EALfbxt8vwGX5k5b
linkToPdf http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMw1V3LbtQwFL0qU4RgAbSAGCjgRRGwiJrYTuIsEJoyDIxKR1kAalfGie1qFk1KkimaX-Pr8M2jzAZ2XbBNrDyPz71-3HMA9rl1tKeV8JTPtMczrT2hMutpHTETGxpRrluziXixECcnSboFv4ZaGNxWOXBiS9S6zHGO_MABj4mY0ZAd2H5bRDqdvbv44aGDFK60DnYaHUSOzPqnG77Vb-dT969fUjr78OX9J693GPByFrHGi5OQ81hxQ_1MZ5kNVehTzRMllKIiyAOmotzmnCcR1yy0GLyFax0nWviZzZi77g3Ydik55yPYTufH6enVDA8qborA74oCGUt8XJPGABtxNJXfCIOtW8DfYkIb6Gb3_udPdB_u9uk1mXT9YQe2TLELdzZEF3dhp6ezmrzuNbffPIBv03Whzpc5WZTakLRaXqp8TTA_XlWGTN0YfYWly2fkIwp8k8mqKVEBVJuKHLo8QJOyIJOm6faOkmOD9dTL-vwhfL2Wt30Eo6IszGMgymhcOPcjRQ2PNE1cLm0D63PlWNIoNob9AQHyolMWkW5EhkCRG0AZwyGi46oJyoG3B8rqTPbsIhPOlOKh5W1-qxPBle_uT4NcG8qMGcMrxJZE0moqlau-9sI9Kcp_yYngKMUmRDiGvQFbsmezWv4B1pN_n34Btxzi5Of54ugp3KZoldxOWO3BqKlW5hnczC-bZV097_sKge_Xjb3fOrpXHA
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Dynamic+Node+Privacy+Feature+Decoupling+Graph+Autoencoder+Based+on+Attention+Mechanism&rft.jtitle=Applied+sciences&rft.au=Huang+Yikai&rft.au=Tang+Jinchuan&rft.au=Dang+Shuping&rft.date=2025-06-01&rft.pub=MDPI+AG&rft.eissn=2076-3417&rft.volume=15&rft.issue=12&rft.spage=6489&rft_id=info:doi/10.3390%2Fapp15126489&rft.externalDBID=HAS_PDF_LINK
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2076-3417&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2076-3417&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2076-3417&client=summon