Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387560482> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W4387560482 abstract "With the growing privacy concerns in recommender systems, recommendation unlearning, i.e., forgetting the impact of specific learned targets, is getting increasing attention. Existing studies predominantly use training data, i.e., model inputs, as the unlearning target. However, we find that attackers can extract private information, i.e., gender, race, and age, from a trained model even if it has not been explicitly encountered during training. We name this unseen information as attribute and treat it as the unlearning target. To protect the sensitive attribute of users, Attribute Unlearning (AU) aims to degrade attacking performance and make target attributes indistinguishable. In this paper, we focus on a strict but practical setting of AU, namely Post-Training Attribute Unlearning (PoT-AU), where unlearning can only be performed after the training of the recommendation model is completed. To address the PoT-AU problem in recommender systems, we design a two-component loss function that consists of i) distinguishability loss: making attribute labels indistinguishable from attackers, and ii) regularization loss: preventing drastic changes in the model that result in a negative impact on recommendation performance. Specifically, we investigate two types of distinguishability measurements, i.e., user-to-user and distribution-to-distribution. We use the stochastic gradient descent algorithm to optimize our proposed loss. Extensive experiments on three real-world datasets demonstrate the effectiveness of our proposed methods." @default.
- W4387560482 created "2023-10-12" @default.
- W4387560482 creator A5028791879 @default.
- W4387560482 creator A5041086326 @default.
- W4387560482 creator A5044163944 @default.
- W4387560482 creator A5058559904 @default.
- W4387560482 creator A5060065671 @default.
- W4387560482 creator A5074603286 @default.
- W4387560482 creator A5091408462 @default.
- W4387560482 date "2023-10-06" @default.
- W4387560482 modified "2023-10-18" @default.
- W4387560482 title "Making Users Indistinguishable: Attribute-wise Unlearning in Recommender Systems" @default.
- W4387560482 doi "https://doi.org/10.1145/3581783.3612418" @default.
- W4387560482 hasPublicationYear "2023" @default.
- W4387560482 type Work @default.
- W4387560482 citedByCount "0" @default.
- W4387560482 crossrefType "posted-content" @default.
- W4387560482 hasAuthorship W4387560482A5028791879 @default.
- W4387560482 hasAuthorship W4387560482A5041086326 @default.
- W4387560482 hasAuthorship W4387560482A5044163944 @default.
- W4387560482 hasAuthorship W4387560482A5058559904 @default.
- W4387560482 hasAuthorship W4387560482A5060065671 @default.
- W4387560482 hasAuthorship W4387560482A5074603286 @default.
- W4387560482 hasAuthorship W4387560482A5091408462 @default.
- W4387560482 hasBestOaLocation W43875604821 @default.
- W4387560482 hasConcept C119857082 @default.
- W4387560482 hasConcept C138885662 @default.
- W4387560482 hasConcept C153258448 @default.
- W4387560482 hasConcept C154945302 @default.
- W4387560482 hasConcept C206688291 @default.
- W4387560482 hasConcept C2776135515 @default.
- W4387560482 hasConcept C2988416141 @default.
- W4387560482 hasConcept C38652104 @default.
- W4387560482 hasConcept C41008148 @default.
- W4387560482 hasConcept C41895202 @default.
- W4387560482 hasConcept C50644808 @default.
- W4387560482 hasConcept C557471498 @default.
- W4387560482 hasConcept C7149132 @default.
- W4387560482 hasConcept C99221444 @default.
- W4387560482 hasConceptScore W4387560482C119857082 @default.
- W4387560482 hasConceptScore W4387560482C138885662 @default.
- W4387560482 hasConceptScore W4387560482C153258448 @default.
- W4387560482 hasConceptScore W4387560482C154945302 @default.
- W4387560482 hasConceptScore W4387560482C206688291 @default.
- W4387560482 hasConceptScore W4387560482C2776135515 @default.
- W4387560482 hasConceptScore W4387560482C2988416141 @default.
- W4387560482 hasConceptScore W4387560482C38652104 @default.
- W4387560482 hasConceptScore W4387560482C41008148 @default.
- W4387560482 hasConceptScore W4387560482C41895202 @default.
- W4387560482 hasConceptScore W4387560482C50644808 @default.
- W4387560482 hasConceptScore W4387560482C557471498 @default.
- W4387560482 hasConceptScore W4387560482C7149132 @default.
- W4387560482 hasConceptScore W4387560482C99221444 @default.
- W4387560482 hasLocation W43875604821 @default.
- W4387560482 hasOpenAccess W4387560482 @default.
- W4387560482 hasPrimaryLocation W43875604821 @default.
- W4387560482 hasRelatedWork W2754816816 @default.
- W4387560482 hasRelatedWork W2895097035 @default.
- W4387560482 hasRelatedWork W3008318776 @default.
- W4387560482 hasRelatedWork W3020853991 @default.
- W4387560482 hasRelatedWork W3160167280 @default.
- W4387560482 hasRelatedWork W3171021120 @default.
- W4387560482 hasRelatedWork W4206903459 @default.
- W4387560482 hasRelatedWork W4231621013 @default.
- W4387560482 hasRelatedWork W4362706668 @default.
- W4387560482 hasRelatedWork W4366280654 @default.
- W4387560482 isParatext "false" @default.
- W4387560482 isRetracted "false" @default.
- W4387560482 workType "article" @default.