Matches in SemOpenAlex for { <https://semopenalex.org/work/W3172284381> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W3172284381 abstract "Word embeddings are widely used in Natural Language Processing (NLP) for a vast range of applications. However, it has been consistently proven that these embeddings reflect the same human biases that exist in the data used to train them. Most of the introduced bias indicators to reveal word embeddings’ bias are average-based indicators based on the cosine similarity measure. In this study, we examine the impacts of different similarity measures as well as other descriptive techniques than averaging in measuring the biases of contextual and non-contextual word embeddings. We show that the extent of revealed biases in word embeddings depends on the descriptive statistics and similarity measures used to measure the bias. We found that over the ten categories of word embedding association tests, Mahalanobis distance reveals the smallest bias, and Euclidean distance reveals the largest bias in word embeddings. In addition, the contextual models reveal less severe biases than the non-contextual word embedding models." @default.
- W3172284381 created "2021-06-22" @default.
- W3172284381 creator A5084826418 @default.
- W3172284381 creator A5085722495 @default.
- W3172284381 date "2021-01-01" @default.
- W3172284381 modified "2023-09-25" @default.
- W3172284381 title "Measuring Biases of Word Embeddings: What Similarity Measures and Descriptive Statistics to Use?" @default.
- W3172284381 cites W1981398747 @default.
- W3172284381 cites W1982186270 @default.
- W3172284381 cites W2066255970 @default.
- W3172284381 cites W2107031757 @default.
- W3172284381 cites W2140534852 @default.
- W3172284381 cites W2145574511 @default.
- W3172284381 cites W2153579005 @default.
- W3172284381 cites W2250539671 @default.
- W3172284381 cites W2483215953 @default.
- W3172284381 cites W2769358515 @default.
- W3172284381 cites W2893425640 @default.
- W3172284381 cites W2905678540 @default.
- W3172284381 cites W2962739339 @default.
- W3172284381 cites W2963078909 @default.
- W3172284381 cites W2963341956 @default.
- W3172284381 cites W2963723885 @default.
- W3172284381 cites W2972413484 @default.
- W3172284381 cites W2972668795 @default.
- W3172284381 cites W3035241006 @default.
- W3172284381 cites W3035591180 @default.
- W3172284381 cites W3037697022 @default.
- W3172284381 cites W3123223998 @default.
- W3172284381 doi "https://doi.org/10.18653/v1/2021.trustnlp-1.2" @default.
- W3172284381 hasPublicationYear "2021" @default.
- W3172284381 type Work @default.
- W3172284381 sameAs 3172284381 @default.
- W3172284381 citedByCount "0" @default.
- W3172284381 crossrefType "proceedings-article" @default.
- W3172284381 hasAuthorship W3172284381A5084826418 @default.
- W3172284381 hasAuthorship W3172284381A5085722495 @default.
- W3172284381 hasBestOaLocation W31722843811 @default.
- W3172284381 hasConcept C103278499 @default.
- W3172284381 hasConcept C105795698 @default.
- W3172284381 hasConcept C115961682 @default.
- W3172284381 hasConcept C153180895 @default.
- W3172284381 hasConcept C154945302 @default.
- W3172284381 hasConcept C159985019 @default.
- W3172284381 hasConcept C1921717 @default.
- W3172284381 hasConcept C192562407 @default.
- W3172284381 hasConcept C204321447 @default.
- W3172284381 hasConcept C204323151 @default.
- W3172284381 hasConcept C2524010 @default.
- W3172284381 hasConcept C2777462759 @default.
- W3172284381 hasConcept C2780762811 @default.
- W3172284381 hasConcept C33923547 @default.
- W3172284381 hasConcept C41008148 @default.
- W3172284381 hasConcept C41608201 @default.
- W3172284381 hasConcept C90805587 @default.
- W3172284381 hasConceptScore W3172284381C103278499 @default.
- W3172284381 hasConceptScore W3172284381C105795698 @default.
- W3172284381 hasConceptScore W3172284381C115961682 @default.
- W3172284381 hasConceptScore W3172284381C153180895 @default.
- W3172284381 hasConceptScore W3172284381C154945302 @default.
- W3172284381 hasConceptScore W3172284381C159985019 @default.
- W3172284381 hasConceptScore W3172284381C1921717 @default.
- W3172284381 hasConceptScore W3172284381C192562407 @default.
- W3172284381 hasConceptScore W3172284381C204321447 @default.
- W3172284381 hasConceptScore W3172284381C204323151 @default.
- W3172284381 hasConceptScore W3172284381C2524010 @default.
- W3172284381 hasConceptScore W3172284381C2777462759 @default.
- W3172284381 hasConceptScore W3172284381C2780762811 @default.
- W3172284381 hasConceptScore W3172284381C33923547 @default.
- W3172284381 hasConceptScore W3172284381C41008148 @default.
- W3172284381 hasConceptScore W3172284381C41608201 @default.
- W3172284381 hasConceptScore W3172284381C90805587 @default.
- W3172284381 hasLocation W31722843811 @default.
- W3172284381 hasOpenAccess W3172284381 @default.
- W3172284381 hasPrimaryLocation W31722843811 @default.
- W3172284381 hasRelatedWork W2786259169 @default.
- W3172284381 hasRelatedWork W2935713639 @default.
- W3172284381 hasRelatedWork W3083244731 @default.
- W3172284381 hasRelatedWork W3086845375 @default.
- W3172284381 hasRelatedWork W3113012686 @default.
- W3172284381 hasRelatedWork W3114046961 @default.
- W3172284381 hasRelatedWork W3158961393 @default.
- W3172284381 hasRelatedWork W3172284381 @default.
- W3172284381 hasRelatedWork W4286799877 @default.
- W3172284381 hasRelatedWork W4323981018 @default.
- W3172284381 isParatext "false" @default.
- W3172284381 isRetracted "false" @default.
- W3172284381 magId "3172284381" @default.
- W3172284381 workType "article" @default.