Matches in SemOpenAlex for { <https://semopenalex.org/work/W4284672647> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W4284672647 abstract "Societal biases can influence Information Retrieval system results, and conversely, search results can potentially reinforce existing societal biases. Recent research has therefore focused on developing methods for quantifying and mitigating bias in search results and applied them to contemporary retrieval systems that leverage transformer-based language models. In the present work, we expand this direction of research by considering bias mitigation within a framework for contextual document embedding reranking. In this framework, the transformer-based query encoder is optimized for relevance ranking through a list-wise objective, by jointly scoring for the same query a large set of candidate document embeddings in the context of one another, instead of in isolation. At the same time, we impose a regularization loss which penalizes highly scoring documents that deviate from neutrality with respect to a protected attribute (e.g., gender). Our approach for bias mitigation is end-to-end differentiable and efficient. Compared to the existing alternatives for deep neural retrieval architectures, which are based on adversarial training, we demonstrate that it can attain much stronger bias mitigation/fairness. At the same time, for the same amount of bias mitigation, it offers significantly better relevance performance (utility). Crucially, our method allows for a more finely controllable and predictable intensity of bias mitigation, which is essential for practical deployment in production systems." @default.
- W4284672647 created "2022-07-08" @default.
- W4284672647 creator A5008481063 @default.
- W4284672647 creator A5014921416 @default.
- W4284672647 creator A5017877491 @default.
- W4284672647 creator A5082506991 @default.
- W4284672647 date "2022-07-06" @default.
- W4284672647 modified "2023-10-14" @default.
- W4284672647 title "Mitigating Bias in Search Results Through Contextual Document Reranking and Neutrality Regularization" @default.
- W4284672647 cites W1583837637 @default.
- W4284672647 cites W1885634126 @default.
- W4284672647 cites W2006187024 @default.
- W4284672647 cites W2099865247 @default.
- W4284672647 cites W2149252982 @default.
- W4284672647 cites W2591615201 @default.
- W4284672647 cites W2740321901 @default.
- W4284672647 cites W2787991113 @default.
- W4284672647 cites W2796402109 @default.
- W4284672647 cites W3012903288 @default.
- W4284672647 cites W3023202929 @default.
- W4284672647 cites W3028722847 @default.
- W4284672647 cites W3099700870 @default.
- W4284672647 cites W3102961637 @default.
- W4284672647 cites W3125913709 @default.
- W4284672647 cites W3154670582 @default.
- W4284672647 cites W3155480849 @default.
- W4284672647 cites W3155690528 @default.
- W4284672647 cites W3155895380 @default.
- W4284672647 cites W3156988171 @default.
- W4284672647 cites W3157758108 @default.
- W4284672647 cites W3157975787 @default.
- W4284672647 cites W3173172270 @default.
- W4284672647 cites W3180044136 @default.
- W4284672647 cites W3182104952 @default.
- W4284672647 cites W3198073108 @default.
- W4284672647 cites W4281885833 @default.
- W4284672647 cites W4285143031 @default.
- W4284672647 cites W4300482433 @default.
- W4284672647 doi "https://doi.org/10.1145/3477495.3531891" @default.
- W4284672647 hasPublicationYear "2022" @default.
- W4284672647 type Work @default.
- W4284672647 citedByCount "5" @default.
- W4284672647 countsByYear W42846726472022 @default.
- W4284672647 countsByYear W42846726472023 @default.
- W4284672647 crossrefType "proceedings-article" @default.
- W4284672647 hasAuthorship W4284672647A5008481063 @default.
- W4284672647 hasAuthorship W4284672647A5014921416 @default.
- W4284672647 hasAuthorship W4284672647A5017877491 @default.
- W4284672647 hasAuthorship W4284672647A5082506991 @default.
- W4284672647 hasConcept C119857082 @default.
- W4284672647 hasConcept C124101348 @default.
- W4284672647 hasConcept C153083717 @default.
- W4284672647 hasConcept C154945302 @default.
- W4284672647 hasConcept C158154518 @default.
- W4284672647 hasConcept C17744445 @default.
- W4284672647 hasConcept C199539241 @default.
- W4284672647 hasConcept C23123220 @default.
- W4284672647 hasConcept C2776135515 @default.
- W4284672647 hasConcept C37736160 @default.
- W4284672647 hasConcept C41008148 @default.
- W4284672647 hasConceptScore W4284672647C119857082 @default.
- W4284672647 hasConceptScore W4284672647C124101348 @default.
- W4284672647 hasConceptScore W4284672647C153083717 @default.
- W4284672647 hasConceptScore W4284672647C154945302 @default.
- W4284672647 hasConceptScore W4284672647C158154518 @default.
- W4284672647 hasConceptScore W4284672647C17744445 @default.
- W4284672647 hasConceptScore W4284672647C199539241 @default.
- W4284672647 hasConceptScore W4284672647C23123220 @default.
- W4284672647 hasConceptScore W4284672647C2776135515 @default.
- W4284672647 hasConceptScore W4284672647C37736160 @default.
- W4284672647 hasConceptScore W4284672647C41008148 @default.
- W4284672647 hasFunder F4320306076 @default.
- W4284672647 hasLocation W42846726471 @default.
- W4284672647 hasOpenAccess W4284672647 @default.
- W4284672647 hasPrimaryLocation W42846726471 @default.
- W4284672647 hasRelatedWork W1843462531 @default.
- W4284672647 hasRelatedWork W1967370444 @default.
- W4284672647 hasRelatedWork W2294590153 @default.
- W4284672647 hasRelatedWork W2891261571 @default.
- W4284672647 hasRelatedWork W2903917280 @default.
- W4284672647 hasRelatedWork W2961085424 @default.
- W4284672647 hasRelatedWork W3207081511 @default.
- W4284672647 hasRelatedWork W4306674287 @default.
- W4284672647 hasRelatedWork W4307930337 @default.
- W4284672647 hasRelatedWork W4224009465 @default.
- W4284672647 isParatext "false" @default.
- W4284672647 isRetracted "false" @default.
- W4284672647 workType "article" @default.