FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4376132596> ?p ?o ?g. }

• W4376132596 endingPage "17" @default.
• W4376132596 startingPage "1" @default.
• W4376132596 abstract "Recent studies have shown that recommender systems are vulnerable, and it is easy for attackers to inject well-designed malicious profiles into the system, resulting in biased recommendations. We cannot deprive these data's injection right and deny their existence's rationality, making it imperative to study recommendation robustness. Despite impressive emerging work, threat assessment of the bi-level poisoning problem and the imperceptibility of poisoning users remain key challenges to be solved. To this end, we propose Infmix, an efficient poisoning attack strategy. Specifically, Infmix consists of an influence-based threat estimator and a user generator, Usermix. First, the influence-based estimator can efficiently evaluate the user's harm to the recommender system without retraining, which is challenging for existing attacks. Second, Usermix, a distribution-agnostic generator, can generate unnoticeable fake data even with a few known users. Under the guidance of the threat estimator, Infmix can select the users with large attacking impacts from the quasi-real candidates generated by Usermix. Extensive experiments demonstrate Infmix's superiority by attacking six recommendation systems with four real datasets. Additionally, we propose a novel defense strategy, adversarial poisoning training (APT). It mimics the poisoning process by injecting fake users (ERM users) committed to minimizing empirical risk to build a robust system. Similar to Infmix, we also utilize the influence function to solve the bi-level optimization challenge of generating ERM users. Although the idea of “fighting fire with fire” in APT seems counterintuitive, we prove its effectiveness in improving recommendation robustness through theoretical analysis and empirical experiments." @default.
• W4376132596 created "2023-05-12" @default.
• W4376132596 creator A5033069158 @default.
• W4376132596 creator A5048237545 @default.
• W4376132596 creator A5048511473 @default.
• W4376132596 creator A5050609316 @default.
• W4376132596 creator A5085254654 @default.
• W4376132596 date "2023-01-01" @default.
• W4376132596 modified "2023-10-15" @default.
• W4376132596 title "Influence-Driven Data Poisoning for Robust Recommender Systems" @default.
• W4376132596 cites W1987431925 @default.
• W4376132596 cites W1989279326 @default.
• W4376132596 cites W2019313057 @default.
• W4376132596 cites W2022613648 @default.
• W4376132596 cites W2029424893 @default.
• W4376132596 cites W2048672094 @default.
• W4376132596 cites W2054141820 @default.
• W4376132596 cites W2069480896 @default.
• W4376132596 cites W2072651985 @default.
• W4376132596 cites W2110096996 @default.
• W4376132596 cites W2117017885 @default.
• W4376132596 cites W2133699849 @default.
• W4376132596 cites W2152208379 @default.
• W4376132596 cites W2463645429 @default.
• W4376132596 cites W2604662567 @default.
• W4376132596 cites W2605350416 @default.
• W4376132596 cites W2613314732 @default.
• W4376132596 cites W2614794251 @default.
• W4376132596 cites W2794121917 @default.
• W4376132596 cites W2798868970 @default.
• W4376132596 cites W2799032844 @default.
• W4376132596 cites W2801109448 @default.
• W4376132596 cites W2890884621 @default.
• W4376132596 cites W2892160417 @default.
• W4376132596 cites W2897212375 @default.
• W4376132596 cites W2903574258 @default.
• W4376132596 cites W2903665338 @default.
• W4376132596 cites W2913696151 @default.
• W4376132596 cites W2954667104 @default.
• W4376132596 cites W2962907114 @default.
• W4376132596 cites W2972630152 @default.
• W4376132596 cites W2972646741 @default.
• W4376132596 cites W2972965281 @default.
• W4376132596 cites W2997340900 @default.
• W4376132596 cites W3012794253 @default.
• W4376132596 cites W3012798016 @default.
• W4376132596 cites W3031339255 @default.
• W4376132596 cites W3034339445 @default.
• W4376132596 cites W3035298482 @default.
• W4376132596 cites W3048511744 @default.
• W4376132596 cites W3080292238 @default.
• W4376132596 cites W3103801215 @default.
• W4376132596 cites W3104956872 @default.
• W4376132596 cites W3119520312 @default.
• W4376132596 cites W3133578111 @default.
• W4376132596 cites W3133706083 @default.
• W4376132596 cites W3156766660 @default.
• W4376132596 cites W3170976035 @default.
• W4376132596 cites W3202472958 @default.
• W4376132596 doi "https://doi.org/10.1109/tpami.2023.3274759" @default.
• W4376132596 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37163407" @default.
• W4376132596 hasPublicationYear "2023" @default.
• W4376132596 type Work @default.
• W4376132596 citedByCount "1" @default.
• W4376132596 countsByYear W43761325962023 @default.
• W4376132596 crossrefType "journal-article" @default.
• W4376132596 hasAuthorship W4376132596A5033069158 @default.
• W4376132596 hasAuthorship W4376132596A5048237545 @default.
• W4376132596 hasAuthorship W4376132596A5048511473 @default.
• W4376132596 hasAuthorship W4376132596A5050609316 @default.
• W4376132596 hasAuthorship W4376132596A5085254654 @default.
• W4376132596 hasConcept C101097943 @default.
• W4376132596 hasConcept C104317684 @default.
• W4376132596 hasConcept C111472728 @default.
• W4376132596 hasConcept C119857082 @default.
• W4376132596 hasConcept C120936955 @default.
• W4376132596 hasConcept C121332964 @default.
• W4376132596 hasConcept C138885662 @default.
• W4376132596 hasConcept C154945302 @default.
• W4376132596 hasConcept C163258240 @default.
• W4376132596 hasConcept C17744445 @default.
• W4376132596 hasConcept C185592680 @default.
• W4376132596 hasConcept C199539241 @default.
• W4376132596 hasConcept C2777363581 @default.
• W4376132596 hasConcept C2780992000 @default.
• W4376132596 hasConcept C37736160 @default.
• W4376132596 hasConcept C38652104 @default.
• W4376132596 hasConcept C41008148 @default.
• W4376132596 hasConcept C55493867 @default.
• W4376132596 hasConcept C557471498 @default.
• W4376132596 hasConcept C62520636 @default.
• W4376132596 hasConcept C63479239 @default.
• W4376132596 hasConceptScore W4376132596C101097943 @default.
• W4376132596 hasConceptScore W4376132596C104317684 @default.
• W4376132596 hasConceptScore W4376132596C111472728 @default.
• W4376132596 hasConceptScore W4376132596C119857082 @default.
• W4376132596 hasConceptScore W4376132596C120936955 @default.
• W4376132596 hasConceptScore W4376132596C121332964 @default.

• next