Matches in SemOpenAlex for { <https://semopenalex.org/work/W3100283070> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W3100283070 abstract "We present a novel way of injecting factual knowledge about entities into the pretrained BERT model (Devlin et al., 2019): We align Wikipedia2Vec entity vectors (Yamada et al., 2016) with BERT’s native wordpiece vector space and use the aligned entity vectors as if they were wordpiece vectors. The resulting entity-enhanced version of BERT (called E-BERT) is similar in spirit to ERNIE (Zhang et al., 2019) and KnowBert (Peters et al., 2019), but it requires no expensive further pre-training of the BERT encoder. We evaluate E-BERT on unsupervised question answering (QA), supervised relation classification (RC) and entity linking (EL). On all three tasks, E-BERT outperforms BERT and other baselines. We also show quantitatively that the original BERT model is overly reliant on the surface form of entity names (e.g., guessing that someone with an Italian-sounding name speaks Italian), and that E-BERT mitigates this problem." @default.
- W3100283070 created "2020-11-23" @default.
- W3100283070 creator A5042212425 @default.
- W3100283070 creator A5059629985 @default.
- W3100283070 creator A5071144367 @default.
- W3100283070 date "2020-01-01" @default.
- W3100283070 modified "2023-10-03" @default.
- W3100283070 title "E-BERT: Efficient-Yet-Effective Entity Embeddings for BERT" @default.
- W3100283070 cites W11298561 @default.
- W3100283070 cites W1532325895 @default.
- W3100283070 cites W2126725946 @default.
- W3100283070 cites W2594021297 @default.
- W3100283070 cites W2785611959 @default.
- W3100283070 cites W2888236192 @default.
- W3100283070 cites W2890961898 @default.
- W3100283070 cites W2950577311 @default.
- W3100283070 cites W2953356739 @default.
- W3100283070 cites W2963341956 @default.
- W3100283070 cites W2963403868 @default.
- W3100283070 cites W2963855739 @default.
- W3100283070 cites W2964059756 @default.
- W3100283070 cites W2964121744 @default.
- W3100283070 cites W2965373594 @default.
- W3100283070 cites W2966989210 @default.
- W3100283070 cites W2970161131 @default.
- W3100283070 cites W2970283086 @default.
- W3100283070 cites W2970597249 @default.
- W3100283070 cites W2970986510 @default.
- W3100283070 cites W2971869958 @default.
- W3100283070 cites W2983577274 @default.
- W3100283070 cites W2984147501 @default.
- W3100283070 cites W2988975212 @default.
- W3100283070 cites W3005441132 @default.
- W3100283070 cites W3035488502 @default.
- W3100283070 cites W3151929433 @default.
- W3100283070 doi "https://doi.org/10.18653/v1/2020.findings-emnlp.71" @default.
- W3100283070 hasPublicationYear "2020" @default.
- W3100283070 type Work @default.
- W3100283070 sameAs 3100283070 @default.
- W3100283070 citedByCount "61" @default.
- W3100283070 countsByYear W31002830702019 @default.
- W3100283070 countsByYear W31002830702020 @default.
- W3100283070 countsByYear W31002830702021 @default.
- W3100283070 countsByYear W31002830702022 @default.
- W3100283070 countsByYear W31002830702023 @default.
- W3100283070 crossrefType "proceedings-article" @default.
- W3100283070 hasAuthorship W3100283070A5042212425 @default.
- W3100283070 hasAuthorship W3100283070A5059629985 @default.
- W3100283070 hasAuthorship W3100283070A5071144367 @default.
- W3100283070 hasBestOaLocation W31002830702 @default.
- W3100283070 hasConcept C111919701 @default.
- W3100283070 hasConcept C118505674 @default.
- W3100283070 hasConcept C124101348 @default.
- W3100283070 hasConcept C154945302 @default.
- W3100283070 hasConcept C166957645 @default.
- W3100283070 hasConcept C191935318 @default.
- W3100283070 hasConcept C204321447 @default.
- W3100283070 hasConcept C23123220 @default.
- W3100283070 hasConcept C25343380 @default.
- W3100283070 hasConcept C2777045944 @default.
- W3100283070 hasConcept C2778572836 @default.
- W3100283070 hasConcept C41008148 @default.
- W3100283070 hasConcept C95457728 @default.
- W3100283070 hasConceptScore W3100283070C111919701 @default.
- W3100283070 hasConceptScore W3100283070C118505674 @default.
- W3100283070 hasConceptScore W3100283070C124101348 @default.
- W3100283070 hasConceptScore W3100283070C154945302 @default.
- W3100283070 hasConceptScore W3100283070C166957645 @default.
- W3100283070 hasConceptScore W3100283070C191935318 @default.
- W3100283070 hasConceptScore W3100283070C204321447 @default.
- W3100283070 hasConceptScore W3100283070C23123220 @default.
- W3100283070 hasConceptScore W3100283070C25343380 @default.
- W3100283070 hasConceptScore W3100283070C2777045944 @default.
- W3100283070 hasConceptScore W3100283070C2778572836 @default.
- W3100283070 hasConceptScore W3100283070C41008148 @default.
- W3100283070 hasConceptScore W3100283070C95457728 @default.
- W3100283070 hasLocation W31002830701 @default.
- W3100283070 hasLocation W31002830702 @default.
- W3100283070 hasOpenAccess W3100283070 @default.
- W3100283070 hasPrimaryLocation W31002830701 @default.
- W3100283070 hasRelatedWork W10803944 @default.
- W3100283070 hasRelatedWork W14090282 @default.
- W3100283070 hasRelatedWork W14104432 @default.
- W3100283070 hasRelatedWork W2060686 @default.
- W3100283070 hasRelatedWork W232572 @default.
- W3100283070 hasRelatedWork W2800105 @default.
- W3100283070 hasRelatedWork W3884061 @default.
- W3100283070 hasRelatedWork W4531093 @default.
- W3100283070 hasRelatedWork W5470734 @default.
- W3100283070 hasRelatedWork W855327 @default.
- W3100283070 isParatext "false" @default.
- W3100283070 isRetracted "false" @default.
- W3100283070 magId "3100283070" @default.
- W3100283070 workType "article" @default.