FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4284669007> ?p ?o ?g. }

• W4284669007 abstract "Recommender system usually faces popularity bias. From the popularity distribution shift perspective, the normal paradigm trained on exposed items (most are hot items) identifies that recommending popular items more frequently can achieve lower loss, thus injecting popularity information into item property embedding, e.g., id embedding. From the long-tail distribution shift perspective, the sparse interactions of long-tail items lead to insufficient learning of them. The resultant distribution discrepancy between hot and long-tail items would not only inherit the bias, but also amplify the bias. Existing work addresses this issue with inverse propensity scoring (IPS) or causal embeddings. However, we argue that not all popularity biases mean bad effects, i.e., some items show higher popularity due to better quality or conform to current trends, which deserve more recommendations. Blindly seeking unbiased learning may inhibit high-quality or fashionable items. To make better use of the popularity bias, we propose a co-training disentangled domain adaptation network (CD$^2$AN), which can co-train both biased and unbiased models. Specifically, for popularity distribution shift, CD$^2$AN disentangles item property representation and popularity representation from item property embedding. For long-tail distribution shift, we introduce additional unexposed items (most are long-tail items) to align the distribution of hot and long-tail item property representations. Further, from the instances perspective, we carefully design the item similarity regularization to learn comprehensive item representation, which encourages item pairs with more effective co-occurrences patterns to have more similar item property representations. Based on offline evaluations and online A/B tests, we show that CD$^2$AN outperforms the existing debiased solutions. Currently, CD$^2$AN has been successfully deployed at Mobile Taobao App and handling major online traffic." @default.
• W4284669007 created "2022-07-08" @default.
• W4284669007 creator A5012687475 @default.
• W4284669007 creator A5041242808 @default.
• W4284669007 creator A5065789784 @default.
• W4284669007 creator A5069813415 @default.
• W4284669007 creator A5074005818 @default.
• W4284669007 creator A5084417688 @default.
• W4284669007 date "2022-07-06" @default.
• W4284669007 modified "2023-09-30" @default.
• W4284669007 title "Co-training Disentangled Domain Adaptation Network for Leveraging Popularity Bias in Recommenders" @default.
• W4284669007 cites W1886704267 @default.
• W4284669007 cites W2028988057 @default.
• W4284669007 cites W2042281163 @default.
• W4284669007 cites W2142144955 @default.
• W4284669007 cites W2150291618 @default.
• W4284669007 cites W2340502990 @default.
• W4284669007 cites W2507134384 @default.
• W4284669007 cites W2512971201 @default.
• W4284669007 cites W2605350416 @default.
• W4284669007 cites W2723293840 @default.
• W4284669007 cites W2913491198 @default.
• W4284669007 cites W2964117661 @default.
• W4284669007 cites W2972801466 @default.
• W4284669007 cites W2994850640 @default.
• W4284669007 cites W2998702515 @default.
• W4284669007 cites W3013235169 @default.
• W4284669007 cites W3023045848 @default.
• W4284669007 cites W3035544567 @default.
• W4284669007 cites W3089238887 @default.
• W4284669007 cites W3098123823 @default.
• W4284669007 cites W3098468692 @default.
• W4284669007 cites W3099942019 @default.
• W4284669007 cites W3102099102 @default.
• W4284669007 cites W3102755086 @default.
• W4284669007 cites W3115418111 @default.
• W4284669007 cites W3156622960 @default.
• W4284669007 cites W3157014581 @default.
• W4284669007 cites W3170713142 @default.
• W4284669007 cites W3208338073 @default.
• W4284669007 cites W3210807131 @default.
• W4284669007 doi "https://doi.org/10.1145/3477495.3531952" @default.
• W4284669007 hasPublicationYear "2022" @default.
• W4284669007 type Work @default.
• W4284669007 citedByCount "5" @default.
• W4284669007 countsByYear W42846690072023 @default.
• W4284669007 crossrefType "proceedings-article" @default.
• W4284669007 hasAuthorship W4284669007A5012687475 @default.
• W4284669007 hasAuthorship W4284669007A5041242808 @default.
• W4284669007 hasAuthorship W4284669007A5065789784 @default.
• W4284669007 hasAuthorship W4284669007A5069813415 @default.
• W4284669007 hasAuthorship W4284669007A5074005818 @default.
• W4284669007 hasAuthorship W4284669007A5084417688 @default.
• W4284669007 hasConcept C105795698 @default.
• W4284669007 hasConcept C111472728 @default.
• W4284669007 hasConcept C119857082 @default.
• W4284669007 hasConcept C12713177 @default.
• W4284669007 hasConcept C138885662 @default.
• W4284669007 hasConcept C139807058 @default.
• W4284669007 hasConcept C15189868 @default.
• W4284669007 hasConcept C154945302 @default.
• W4284669007 hasConcept C15744967 @default.
• W4284669007 hasConcept C169760540 @default.
• W4284669007 hasConcept C17744445 @default.
• W4284669007 hasConcept C189950617 @default.
• W4284669007 hasConcept C199539241 @default.
• W4284669007 hasConcept C2776359362 @default.
• W4284669007 hasConcept C2780586970 @default.
• W4284669007 hasConcept C33923547 @default.
• W4284669007 hasConcept C41008148 @default.
• W4284669007 hasConcept C41608201 @default.
• W4284669007 hasConcept C77805123 @default.
• W4284669007 hasConcept C94625758 @default.
• W4284669007 hasConceptScore W4284669007C105795698 @default.
• W4284669007 hasConceptScore W4284669007C111472728 @default.
• W4284669007 hasConceptScore W4284669007C119857082 @default.
• W4284669007 hasConceptScore W4284669007C12713177 @default.
• W4284669007 hasConceptScore W4284669007C138885662 @default.
• W4284669007 hasConceptScore W4284669007C139807058 @default.
• W4284669007 hasConceptScore W4284669007C15189868 @default.
• W4284669007 hasConceptScore W4284669007C154945302 @default.
• W4284669007 hasConceptScore W4284669007C15744967 @default.
• W4284669007 hasConceptScore W4284669007C169760540 @default.
• W4284669007 hasConceptScore W4284669007C17744445 @default.
• W4284669007 hasConceptScore W4284669007C189950617 @default.
• W4284669007 hasConceptScore W4284669007C199539241 @default.
• W4284669007 hasConceptScore W4284669007C2776359362 @default.
• W4284669007 hasConceptScore W4284669007C2780586970 @default.
• W4284669007 hasConceptScore W4284669007C33923547 @default.
• W4284669007 hasConceptScore W4284669007C41008148 @default.
• W4284669007 hasConceptScore W4284669007C41608201 @default.
• W4284669007 hasConceptScore W4284669007C77805123 @default.
• W4284669007 hasConceptScore W4284669007C94625758 @default.
• W4284669007 hasLocation W42846690071 @default.
• W4284669007 hasOpenAccess W4284669007 @default.
• W4284669007 hasPrimaryLocation W42846690071 @default.
• W4284669007 hasRelatedWork W2104700403 @default.
• W4284669007 hasRelatedWork W2133515697 @default.
• W4284669007 hasRelatedWork W2194567277 @default.
• W4284669007 hasRelatedWork W2370652538 @default.

• next