FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3125649664> ?p ?o ?g. }

• W3125649664 abstract "Word embeddings trained on large corpora have shown to encode high levels of unfair discriminatory gender, racial, religious and ethnic biases. In contrast, human-written dictionaries describe the meanings of words in a concise, objective and an unbiased manner. We propose a method for debiasing pre-trained word embeddings using dictionaries, without requiring access to the original training resources or any knowledge regarding the word embedding algorithms used. Unlike prior work, our proposed method does not require the types of biases to be pre-defined in the form of word lists, and learns the constraints that must be satisfied by unbiased word embeddings automatically from dictionary definitions of the words. Specifically, we learn an encoder to generate a debiased version of an input word embedding such that it (a) retains the semantics of the pre-trained word embeddings, (b) agrees with the unbiased definition of the word according to the dictionary, and (c) remains orthogonal to the vector space spanned by any biased basis vectors in the pre-trained word embedding space. Experimental results on standard benchmark datasets show that the proposed method can accurately remove unfair biases encoded in pre-trained word embeddings, while preserving useful semantics." @default.
• W3125649664 created "2021-02-01" @default.
• W3125649664 creator A5005531754 @default.
• W3125649664 creator A5073503574 @default.
• W3125649664 date "2021-01-23" @default.
• W3125649664 modified "2023-09-26" @default.
• W3125649664 title "Dictionary-based Debiasing of Pre-trained Word Embeddings" @default.
• W3125649664 cites W1503259811 @default.
• W3125649664 cites W1569415500 @default.
• W3125649664 cites W1807381799 @default.
• W3125649664 cites W1832693441 @default.
• W3125649664 cites W1854884267 @default.
• W3125649664 cites W2038721957 @default.
• W3125649664 cites W2067438047 @default.
• W3125649664 cites W2080100102 @default.
• W3125649664 cites W2095705004 @default.
• W3125649664 cites W2137735870 @default.
• W3125649664 cites W2140534852 @default.
• W3125649664 cites W2141599568 @default.
• W3125649664 cites W2142625445 @default.
• W3125649664 cites W2153579005 @default.
• W3125649664 cites W2154455818 @default.
• W3125649664 cites W2250189634 @default.
• W3125649664 cites W2250539671 @default.
• W3125649664 cites W2251012068 @default.
• W3125649664 cites W2483215953 @default.
• W3125649664 cites W2493916176 @default.
• W3125649664 cites W2752172973 @default.
• W3125649664 cites W2758506174 @default.
• W3125649664 cites W2798962680 @default.
• W3125649664 cites W2803176955 @default.
• W3125649664 cites W2888161220 @default.
• W3125649664 cites W2889624842 @default.
• W3125649664 cites W2891958973 @default.
• W3125649664 cites W2893425640 @default.
• W3125649664 cites W2897630418 @default.
• W3125649664 cites W2922274844 @default.
• W3125649664 cites W2926555354 @default.
• W3125649664 cites W2942160782 @default.
• W3125649664 cites W2949969209 @default.
• W3125649664 cites W2950199579 @default.
• W3125649664 cites W2950866572 @default.
• W3125649664 cites W2952349219 @default.
• W3125649664 cites W2953332543 @default.
• W3125649664 cites W2954275542 @default.
• W3125649664 cites W2962739339 @default.
• W3125649664 cites W2962787423 @default.
• W3125649664 cites W2963078909 @default.
• W3125649664 cites W2963341956 @default.
• W3125649664 cites W2963426755 @default.
• W3125649664 cites W2963446520 @default.
• W3125649664 cites W2963457723 @default.
• W3125649664 cites W2963526187 @default.
• W3125649664 cites W2963612262 @default.
• W3125649664 cites W2963659646 @default.
• W3125649664 cites W2963691697 @default.
• W3125649664 cites W2963879260 @default.
• W3125649664 cites W2964121744 @default.
• W3125649664 cites W2964222246 @default.
• W3125649664 cites W2970340602 @default.
• W3125649664 cites W2970800693 @default.
• W3125649664 cites W2971015127 @default.
• W3125649664 cites W2971307358 @default.
• W3125649664 cites W2972972637 @default.
• W3125649664 cites W3035241006 @default.
• W3125649664 cites W3037697022 @default.
• W3125649664 cites W3118009250 @default.
• W3125649664 cites W3122810052 @default.
• W3125649664 doi "https://doi.org/10.48550/arxiv.2101.09525" @default.
• W3125649664 hasPublicationYear "2021" @default.
• W3125649664 type Work @default.
• W3125649664 sameAs 3125649664 @default.
• W3125649664 citedByCount "3" @default.
• W3125649664 countsByYear W31256496642021 @default.
• W3125649664 crossrefType "posted-content" @default.
• W3125649664 hasAuthorship W3125649664A5005531754 @default.
• W3125649664 hasAuthorship W3125649664A5073503574 @default.
• W3125649664 hasBestOaLocation W31256496641 @default.
• W3125649664 hasConcept C104317684 @default.
• W3125649664 hasConcept C111919701 @default.
• W3125649664 hasConcept C118505674 @default.
• W3125649664 hasConcept C125411270 @default.
• W3125649664 hasConcept C130318100 @default.
• W3125649664 hasConcept C13280743 @default.
• W3125649664 hasConcept C138885662 @default.
• W3125649664 hasConcept C154945302 @default.
• W3125649664 hasConcept C15744967 @default.
• W3125649664 hasConcept C184337299 @default.
• W3125649664 hasConcept C185592680 @default.
• W3125649664 hasConcept C185798385 @default.
• W3125649664 hasConcept C188147891 @default.
• W3125649664 hasConcept C199360897 @default.
• W3125649664 hasConcept C204321447 @default.
• W3125649664 hasConcept C205649164 @default.
• W3125649664 hasConcept C2777462759 @default.
• W3125649664 hasConcept C2778572836 @default.
• W3125649664 hasConcept C2778828372 @default.
• W3125649664 hasConcept C2779458634 @default.
• W3125649664 hasConcept C28490314 @default.
• W3125649664 hasConcept C41008148 @default.

• next