Matches in SemOpenAlex for { <https://semopenalex.org/work/W4367046960> ?p ?o ?g. }
Showing items 1 to 98 of
98
with 100 items per page.
- W4367046960 abstract "Knowledge Graph Embedding (KGE) is a fundamental technique that extracts expressive representation from knowledge graph (KG) to facilitate diverse downstream tasks. The emerging federated KGE (FKGE) collaboratively trains from distributed KGs held among clients while avoiding exchanging clients’ sensitive raw KGs, which can still suffer from privacy threats as evidenced in other federated model trainings (e.g., neural networks). However, quantifying and defending against such privacy threats remain unexplored for FKGE which possesses unique properties not shared by previously studied models. In this paper, we conduct the first holistic study of the privacy threat on FKGE from both attack and defense perspectives. For the attack, we quantify the privacy threat by proposing three new inference attacks, which reveal substantial privacy risk by successfully inferring the existence of the KG triple from victim clients. For the defense, we propose DP-Flames, a novel differentially private FKGE with private selection, which offers a better privacy-utility tradeoff by exploiting the entity-binding sparse gradient property of FKGE and comes with a tight privacy accountant by incorporating the state-of-the-art private selection technique. We further propose an adaptive privacy budget allocation policy to dynamically adjust defense magnitude across the training procedure. Comprehensive evaluations demonstrate that the proposed defense can successfully mitigate the privacy threat by effectively reducing the success rate of inference attacks from to on average with only a modest utility decrease." @default.
- W4367046960 created "2023-04-27" @default.
- W4367046960 creator A5005443916 @default.
- W4367046960 creator A5009452976 @default.
- W4367046960 creator A5022202004 @default.
- W4367046960 creator A5038346302 @default.
- W4367046960 creator A5043524348 @default.
- W4367046960 creator A5060380023 @default.
- W4367046960 creator A5064942736 @default.
- W4367046960 creator A5080477405 @default.
- W4367046960 date "2023-04-30" @default.
- W4367046960 modified "2023-09-25" @default.
- W4367046960 title "Quantifying and Defending against Privacy Threats on Federated Knowledge Graph Embedding" @default.
- W4367046960 cites W1894439495 @default.
- W4367046960 cites W2074418712 @default.
- W4367046960 cites W2080133951 @default.
- W4367046960 cites W2094728533 @default.
- W4367046960 cites W2138865266 @default.
- W4367046960 cites W2250635077 @default.
- W4367046960 cites W2473418344 @default.
- W4367046960 cites W2509893387 @default.
- W4367046960 cites W2887995258 @default.
- W4367046960 cites W2912023992 @default.
- W4367046960 cites W2930926105 @default.
- W4367046960 cites W2962886429 @default.
- W4367046960 cites W2963506795 @default.
- W4367046960 cites W3091870957 @default.
- W4367046960 cites W3176539131 @default.
- W4367046960 cites W3196012972 @default.
- W4367046960 cites W3210948155 @default.
- W4367046960 cites W4200304743 @default.
- W4367046960 doi "https://doi.org/10.1145/3543507.3583450" @default.
- W4367046960 hasPublicationYear "2023" @default.
- W4367046960 type Work @default.
- W4367046960 citedByCount "0" @default.
- W4367046960 crossrefType "proceedings-article" @default.
- W4367046960 hasAuthorship W4367046960A5005443916 @default.
- W4367046960 hasAuthorship W4367046960A5009452976 @default.
- W4367046960 hasAuthorship W4367046960A5022202004 @default.
- W4367046960 hasAuthorship W4367046960A5038346302 @default.
- W4367046960 hasAuthorship W4367046960A5043524348 @default.
- W4367046960 hasAuthorship W4367046960A5060380023 @default.
- W4367046960 hasAuthorship W4367046960A5064942736 @default.
- W4367046960 hasAuthorship W4367046960A5080477405 @default.
- W4367046960 hasBestOaLocation W43670469602 @default.
- W4367046960 hasConcept C102938260 @default.
- W4367046960 hasConcept C108827166 @default.
- W4367046960 hasConcept C123201435 @default.
- W4367046960 hasConcept C132525143 @default.
- W4367046960 hasConcept C140547941 @default.
- W4367046960 hasConcept C154945302 @default.
- W4367046960 hasConcept C17744445 @default.
- W4367046960 hasConcept C199539241 @default.
- W4367046960 hasConcept C2776214188 @default.
- W4367046960 hasConcept C2776359362 @default.
- W4367046960 hasConcept C38652104 @default.
- W4367046960 hasConcept C41008148 @default.
- W4367046960 hasConcept C41608201 @default.
- W4367046960 hasConcept C75564084 @default.
- W4367046960 hasConcept C80444323 @default.
- W4367046960 hasConcept C81917197 @default.
- W4367046960 hasConcept C94625758 @default.
- W4367046960 hasConceptScore W4367046960C102938260 @default.
- W4367046960 hasConceptScore W4367046960C108827166 @default.
- W4367046960 hasConceptScore W4367046960C123201435 @default.
- W4367046960 hasConceptScore W4367046960C132525143 @default.
- W4367046960 hasConceptScore W4367046960C140547941 @default.
- W4367046960 hasConceptScore W4367046960C154945302 @default.
- W4367046960 hasConceptScore W4367046960C17744445 @default.
- W4367046960 hasConceptScore W4367046960C199539241 @default.
- W4367046960 hasConceptScore W4367046960C2776214188 @default.
- W4367046960 hasConceptScore W4367046960C2776359362 @default.
- W4367046960 hasConceptScore W4367046960C38652104 @default.
- W4367046960 hasConceptScore W4367046960C41008148 @default.
- W4367046960 hasConceptScore W4367046960C41608201 @default.
- W4367046960 hasConceptScore W4367046960C75564084 @default.
- W4367046960 hasConceptScore W4367046960C80444323 @default.
- W4367046960 hasConceptScore W4367046960C81917197 @default.
- W4367046960 hasConceptScore W4367046960C94625758 @default.
- W4367046960 hasFunder F4320321001 @default.
- W4367046960 hasFunder F4320335777 @default.
- W4367046960 hasLocation W43670469601 @default.
- W4367046960 hasLocation W43670469602 @default.
- W4367046960 hasOpenAccess W4367046960 @default.
- W4367046960 hasPrimaryLocation W43670469601 @default.
- W4367046960 hasRelatedWork W174580382 @default.
- W4367046960 hasRelatedWork W2116878667 @default.
- W4367046960 hasRelatedWork W2132024542 @default.
- W4367046960 hasRelatedWork W2163661494 @default.
- W4367046960 hasRelatedWork W2626007443 @default.
- W4367046960 hasRelatedWork W2963335492 @default.
- W4367046960 hasRelatedWork W3038106605 @default.
- W4367046960 hasRelatedWork W3135360054 @default.
- W4367046960 hasRelatedWork W4241527182 @default.
- W4367046960 hasRelatedWork W857570378 @default.
- W4367046960 isParatext "false" @default.
- W4367046960 isRetracted "false" @default.
- W4367046960 workType "article" @default.