Matches in SemOpenAlex for { <https://semopenalex.org/work/W3211920512> ?p ?o ?g. }
- W3211920512 abstract "The rise of pre-trained language models has yielded substantial progress in the vast majority of Natural Language Processing (NLP) tasks. However, a generic approach towards the pre-training procedure can naturally be sub-optimal in some cases. Particularly, fine-tuning a pre-trained language model on a source domain and then applying it to a different target domain, results in a sharp performance decline of the eventual classifier for many source-target domain pairs. Moreover, in some NLP tasks, the output categories substantially differ between domains, making adaptation even more challenging. This, for example, happens in the task of aspect extraction, where the aspects of interest of reviews of, e.g., restaurants or electronic devices may be very different. This paper presents a new fine-tuning scheme for BERT, which aims to address the above challenges. We name this scheme DILBERT: Domain Invariant Learning with BERT, and customize it for aspect extraction in the unsupervised domain adaptation setting. DILBERT harnesses the categorical information of both the source and the target domains to guide the pre-training process towards a more domain and category invariant representation, thus closing the gap between the domains. We show that DILBERT yields substantial improvements over state-of-the-art baselines while using a fraction of the unlabeled data, particularly in more challenging domain adaptation setups." @default.
- W3211920512 created "2021-11-22" @default.
- W3211920512 creator A5003744048 @default.
- W3211920512 creator A5054952724 @default.
- W3211920512 creator A5063658122 @default.
- W3211920512 date "2021-01-01" @default.
- W3211920512 modified "2023-10-13" @default.
- W3211920512 title "DILBERT: Customized Pre-Training for Domain Adaptation with Category Shift, with an Application to Aspect Extraction" @default.
- W3211920512 cites W1517771839 @default.
- W3211920512 cites W1731081199 @default.
- W3211920512 cites W1859957297 @default.
- W3211920512 cites W2027731328 @default.
- W3211920512 cites W2096873754 @default.
- W3211920512 cites W2097606805 @default.
- W3211920512 cites W2100362224 @default.
- W3211920512 cites W2105523772 @default.
- W3211920512 cites W2120354757 @default.
- W3211920512 cites W2131953535 @default.
- W3211920512 cites W2148034183 @default.
- W3211920512 cites W2158108973 @default.
- W3211920512 cites W2159457224 @default.
- W3211920512 cites W2163302275 @default.
- W3211920512 cites W2215421138 @default.
- W3211920512 cites W2251294039 @default.
- W3211920512 cites W2251648804 @default.
- W3211920512 cites W2252024663 @default.
- W3211920512 cites W2604356472 @default.
- W3211920512 cites W2803777992 @default.
- W3211920512 cites W2889720764 @default.
- W3211920512 cites W2908510526 @default.
- W3211920512 cites W2925618549 @default.
- W3211920512 cites W2953217208 @default.
- W3211920512 cites W2962750142 @default.
- W3211920512 cites W2962843214 @default.
- W3211920512 cites W2963341956 @default.
- W3211920512 cites W2963403868 @default.
- W3211920512 cites W2963463240 @default.
- W3211920512 cites W2963571341 @default.
- W3211920512 cites W2965373594 @default.
- W3211920512 cites W2971014768 @default.
- W3211920512 cites W2982455176 @default.
- W3211920512 cites W3035419191 @default.
- W3211920512 cites W3083515246 @default.
- W3211920512 cites W3090047754 @default.
- W3211920512 cites W3102449459 @default.
- W3211920512 cites W3114610051 @default.
- W3211920512 cites W3117560413 @default.
- W3211920512 doi "https://doi.org/10.18653/v1/2021.emnlp-main.20" @default.
- W3211920512 hasPublicationYear "2021" @default.
- W3211920512 type Work @default.
- W3211920512 sameAs 3211920512 @default.
- W3211920512 citedByCount "7" @default.
- W3211920512 countsByYear W32119205122022 @default.
- W3211920512 countsByYear W32119205122023 @default.
- W3211920512 crossrefType "proceedings-article" @default.
- W3211920512 hasAuthorship W3211920512A5003744048 @default.
- W3211920512 hasAuthorship W3211920512A5054952724 @default.
- W3211920512 hasAuthorship W3211920512A5063658122 @default.
- W3211920512 hasBestOaLocation W32119205121 @default.
- W3211920512 hasConcept C119857082 @default.
- W3211920512 hasConcept C120665830 @default.
- W3211920512 hasConcept C121332964 @default.
- W3211920512 hasConcept C134306372 @default.
- W3211920512 hasConcept C137293760 @default.
- W3211920512 hasConcept C139807058 @default.
- W3211920512 hasConcept C154945302 @default.
- W3211920512 hasConcept C190470478 @default.
- W3211920512 hasConcept C204321447 @default.
- W3211920512 hasConcept C2776434776 @default.
- W3211920512 hasConcept C33923547 @default.
- W3211920512 hasConcept C36503486 @default.
- W3211920512 hasConcept C37914503 @default.
- W3211920512 hasConcept C41008148 @default.
- W3211920512 hasConcept C5274069 @default.
- W3211920512 hasConcept C95623464 @default.
- W3211920512 hasConceptScore W3211920512C119857082 @default.
- W3211920512 hasConceptScore W3211920512C120665830 @default.
- W3211920512 hasConceptScore W3211920512C121332964 @default.
- W3211920512 hasConceptScore W3211920512C134306372 @default.
- W3211920512 hasConceptScore W3211920512C137293760 @default.
- W3211920512 hasConceptScore W3211920512C139807058 @default.
- W3211920512 hasConceptScore W3211920512C154945302 @default.
- W3211920512 hasConceptScore W3211920512C190470478 @default.
- W3211920512 hasConceptScore W3211920512C204321447 @default.
- W3211920512 hasConceptScore W3211920512C2776434776 @default.
- W3211920512 hasConceptScore W3211920512C33923547 @default.
- W3211920512 hasConceptScore W3211920512C36503486 @default.
- W3211920512 hasConceptScore W3211920512C37914503 @default.
- W3211920512 hasConceptScore W3211920512C41008148 @default.
- W3211920512 hasConceptScore W3211920512C5274069 @default.
- W3211920512 hasConceptScore W3211920512C95623464 @default.
- W3211920512 hasLocation W32119205121 @default.
- W3211920512 hasOpenAccess W3211920512 @default.
- W3211920512 hasPrimaryLocation W32119205121 @default.
- W3211920512 hasRelatedWork W3080655457 @default.
- W3211920512 hasRelatedWork W3093803318 @default.
- W3211920512 hasRelatedWork W3136267388 @default.
- W3211920512 hasRelatedWork W3166286441 @default.
- W3211920512 hasRelatedWork W3186065094 @default.
- W3211920512 hasRelatedWork W3186441233 @default.