Matches in SemOpenAlex for { <https://semopenalex.org/work/W4226307069> ?p ?o ?g. }
- W4226307069 endingPage "913" @default.
- W4226307069 startingPage "900" @default.
- W4226307069 abstract "Matrix factorization (MF) approximates unobserved ratings in a rating matrix, whose rows correspond to users and columns correspond to items to be rated, and has been serving as a fundamental building block in recommendation systems. This paper comprehensively studies the problem of matrix factorization in different federated learning (FL) settings, where a set of parties want to cooperate in training but refuse to share data directly. We first propose a generic algorithmic framework for various settings of <u>f</u>ederated <u>m</u>atrix <u>f</u>actorization (FMF) and provide a theoretical convergence guarantee. We then systematically characterize privacy-leakage risks in data collection, training, and publishing stages for three different settings and introduce privacy notions to provide end-to-end privacy protections. The first one is vertical federated learning (VFL), where multiple parties have the ratings from the same set of users but on disjoint sets of items. The second one is horizontal federated learning (HFL), where parties have ratings from different sets of users but on the same set of items. The third setting is local federated learning (LFL), where the ratings of the users are only stored on their local devices. We introduce adapted versions of FMF with the privacy notions guaranteed in the three settings. In particular, a new private learning technique called embedding clipping is introduced and used in all the three settings to ensure differential privacy. For the LFL setting, we combine differential privacy with secure aggregation to protect the communication between user devices and the server with a strength similar to the local differential privacy model, but much better accuracy. We perform experiments to demonstrate the effectiveness of our approaches." @default.
- W4226307069 created "2022-05-05" @default.
- W4226307069 creator A5010853051 @default.
- W4226307069 creator A5031944443 @default.
- W4226307069 creator A5040297543 @default.
- W4226307069 creator A5041232755 @default.
- W4226307069 creator A5057864403 @default.
- W4226307069 date "2021-12-01" @default.
- W4226307069 modified "2023-10-16" @default.
- W4226307069 title "Federated matrix factorization with privacy guarantee" @default.
- W4226307069 cites W1479822238 @default.
- W4226307069 cites W1968752677 @default.
- W4226307069 cites W1980147176 @default.
- W4226307069 cites W1981631259 @default.
- W4226307069 cites W1985511977 @default.
- W4226307069 cites W2009952684 @default.
- W4226307069 cites W2054141820 @default.
- W4226307069 cites W2096870293 @default.
- W4226307069 cites W2110868467 @default.
- W4226307069 cites W2144359569 @default.
- W4226307069 cites W2219888463 @default.
- W4226307069 cites W2395278586 @default.
- W4226307069 cites W2401983063 @default.
- W4226307069 cites W2473418344 @default.
- W4226307069 cites W2520881573 @default.
- W4226307069 cites W2606065148 @default.
- W4226307069 cites W2767079719 @default.
- W4226307069 cites W2789607830 @default.
- W4226307069 cites W2912213068 @default.
- W4226307069 cites W2944951172 @default.
- W4226307069 cites W2951059495 @default.
- W4226307069 cites W2963456518 @default.
- W4226307069 cites W2964341035 @default.
- W4226307069 cites W2970408908 @default.
- W4226307069 cites W3015535799 @default.
- W4226307069 cites W3015807225 @default.
- W4226307069 cites W3016632787 @default.
- W4226307069 cites W3047789363 @default.
- W4226307069 cites W3096328345 @default.
- W4226307069 cites W3101860164 @default.
- W4226307069 cites W3102407811 @default.
- W4226307069 cites W3105446653 @default.
- W4226307069 cites W3113714175 @default.
- W4226307069 cites W4206320562 @default.
- W4226307069 doi "https://doi.org/10.14778/3503585.3503598" @default.
- W4226307069 hasPublicationYear "2021" @default.
- W4226307069 type Work @default.
- W4226307069 citedByCount "6" @default.
- W4226307069 countsByYear W42263070692022 @default.
- W4226307069 countsByYear W42263070692023 @default.
- W4226307069 crossrefType "journal-article" @default.
- W4226307069 hasAuthorship W4226307069A5010853051 @default.
- W4226307069 hasAuthorship W4226307069A5031944443 @default.
- W4226307069 hasAuthorship W4226307069A5040297543 @default.
- W4226307069 hasAuthorship W4226307069A5041232755 @default.
- W4226307069 hasAuthorship W4226307069A5057864403 @default.
- W4226307069 hasConcept C106487976 @default.
- W4226307069 hasConcept C11413529 @default.
- W4226307069 hasConcept C114614502 @default.
- W4226307069 hasConcept C118615104 @default.
- W4226307069 hasConcept C121332964 @default.
- W4226307069 hasConcept C124101348 @default.
- W4226307069 hasConcept C154945302 @default.
- W4226307069 hasConcept C158693339 @default.
- W4226307069 hasConcept C159985019 @default.
- W4226307069 hasConcept C162324750 @default.
- W4226307069 hasConcept C177264268 @default.
- W4226307069 hasConcept C187834632 @default.
- W4226307069 hasConcept C192562407 @default.
- W4226307069 hasConcept C199360897 @default.
- W4226307069 hasConcept C23123220 @default.
- W4226307069 hasConcept C23130292 @default.
- W4226307069 hasConcept C2777210771 @default.
- W4226307069 hasConcept C2777303404 @default.
- W4226307069 hasConcept C2992525071 @default.
- W4226307069 hasConcept C33923547 @default.
- W4226307069 hasConcept C41008148 @default.
- W4226307069 hasConcept C41608201 @default.
- W4226307069 hasConcept C42355184 @default.
- W4226307069 hasConcept C45340560 @default.
- W4226307069 hasConcept C50522688 @default.
- W4226307069 hasConcept C51632099 @default.
- W4226307069 hasConcept C557471498 @default.
- W4226307069 hasConcept C62520636 @default.
- W4226307069 hasConcept C80444323 @default.
- W4226307069 hasConceptScore W4226307069C106487976 @default.
- W4226307069 hasConceptScore W4226307069C11413529 @default.
- W4226307069 hasConceptScore W4226307069C114614502 @default.
- W4226307069 hasConceptScore W4226307069C118615104 @default.
- W4226307069 hasConceptScore W4226307069C121332964 @default.
- W4226307069 hasConceptScore W4226307069C124101348 @default.
- W4226307069 hasConceptScore W4226307069C154945302 @default.
- W4226307069 hasConceptScore W4226307069C158693339 @default.
- W4226307069 hasConceptScore W4226307069C159985019 @default.
- W4226307069 hasConceptScore W4226307069C162324750 @default.
- W4226307069 hasConceptScore W4226307069C177264268 @default.
- W4226307069 hasConceptScore W4226307069C187834632 @default.
- W4226307069 hasConceptScore W4226307069C192562407 @default.