Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386938161> ?p ?o ?g. }
- W4386938161 endingPage "542" @default.
- W4386938161 startingPage "530" @default.
- W4386938161 abstract "Aspect-based sentiment analysis is a challenging yet critical task for recognizing emotions in text, with various applications in social media, commodity reviews, and movie comments. Many researchers are working on developing more powerful sentiment analysis models. Most existing models use the pre-trained language models based fine-tuning paradigm, which only utilizes the encoder parameters of pre-trained language models. However, this approach fails to effectively leverage the prior knowledge revealed in pre-trained language models. To address these issues, we propose a novel approach, Target Word Transferred Language Model for aspect-based sentiment analysis (WordTransABSA), which investigates the potential of the pre-training scheme of pre-trained language models. WordTransABSA is an encoder-decoder architecture built on top of the Masked Language Model of Bidirectional Encoder Representation from Transformers. During the training procedure, we reformulate the previous generic fine-tuning models as a “Masked Language Model” task, which follows the original BERT pre-training paradigm. WordTransABSA takes full advantage of the versatile linguistic knowledge of Pre-trained Language Model, resulting in competitive accuracy compared with recent baselines, especially in data-insufficient scenarios. We have made our code publicly available on GitHub ( https://github.com/albert-jin/WordTransABSA )." @default.
- W4386938161 created "2023-09-22" @default.
- W4386938161 creator A5004322083 @default.
- W4386938161 creator A5023324327 @default.
- W4386938161 creator A5027773843 @default.
- W4386938161 creator A5051652027 @default.
- W4386938161 creator A5076389573 @default.
- W4386938161 date "2023-01-01" @default.
- W4386938161 modified "2023-10-16" @default.
- W4386938161 title "Using Masked Language Modeling to Enhance BERT-Based Aspect-Based Sentiment Analysis for Affective Token Prediction" @default.
- W4386938161 cites W2040467972 @default.
- W4386938161 cites W2251124635 @default.
- W4386938161 cites W2251648804 @default.
- W4386938161 cites W2756381707 @default.
- W4386938161 cites W2925618549 @default.
- W4386938161 cites W2947851192 @default.
- W4386938161 cites W2954278700 @default.
- W4386938161 cites W2963341956 @default.
- W4386938161 cites W2969743835 @default.
- W4386938161 cites W2985056549 @default.
- W4386938161 cites W3034999214 @default.
- W4386938161 cites W3105174597 @default.
- W4386938161 cites W3173777717 @default.
- W4386938161 cites W3176680950 @default.
- W4386938161 cites W3198659451 @default.
- W4386938161 cites W4225794827 @default.
- W4386938161 cites W4289077018 @default.
- W4386938161 cites W4294567976 @default.
- W4386938161 cites W4296708197 @default.
- W4386938161 cites W4297385049 @default.
- W4386938161 cites W4313487628 @default.
- W4386938161 cites W4365802841 @default.
- W4386938161 cites W4386002638 @default.
- W4386938161 doi "https://doi.org/10.1007/978-3-031-44204-9_44" @default.
- W4386938161 hasPublicationYear "2023" @default.
- W4386938161 type Work @default.
- W4386938161 citedByCount "0" @default.
- W4386938161 crossrefType "book-chapter" @default.
- W4386938161 hasAuthorship W4386938161A5004322083 @default.
- W4386938161 hasAuthorship W4386938161A5023324327 @default.
- W4386938161 hasAuthorship W4386938161A5027773843 @default.
- W4386938161 hasAuthorship W4386938161A5051652027 @default.
- W4386938161 hasAuthorship W4386938161A5076389573 @default.
- W4386938161 hasConcept C111919701 @default.
- W4386938161 hasConcept C118505674 @default.
- W4386938161 hasConcept C119857082 @default.
- W4386938161 hasConcept C121332964 @default.
- W4386938161 hasConcept C137293760 @default.
- W4386938161 hasConcept C138885662 @default.
- W4386938161 hasConcept C153083717 @default.
- W4386938161 hasConcept C154945302 @default.
- W4386938161 hasConcept C162324750 @default.
- W4386938161 hasConcept C165801399 @default.
- W4386938161 hasConcept C187736073 @default.
- W4386938161 hasConcept C204321447 @default.
- W4386938161 hasConcept C2780451532 @default.
- W4386938161 hasConcept C38652104 @default.
- W4386938161 hasConcept C41008148 @default.
- W4386938161 hasConcept C41895202 @default.
- W4386938161 hasConcept C48145219 @default.
- W4386938161 hasConcept C62520636 @default.
- W4386938161 hasConcept C66322947 @default.
- W4386938161 hasConcept C66402592 @default.
- W4386938161 hasConcept C90805587 @default.
- W4386938161 hasConceptScore W4386938161C111919701 @default.
- W4386938161 hasConceptScore W4386938161C118505674 @default.
- W4386938161 hasConceptScore W4386938161C119857082 @default.
- W4386938161 hasConceptScore W4386938161C121332964 @default.
- W4386938161 hasConceptScore W4386938161C137293760 @default.
- W4386938161 hasConceptScore W4386938161C138885662 @default.
- W4386938161 hasConceptScore W4386938161C153083717 @default.
- W4386938161 hasConceptScore W4386938161C154945302 @default.
- W4386938161 hasConceptScore W4386938161C162324750 @default.
- W4386938161 hasConceptScore W4386938161C165801399 @default.
- W4386938161 hasConceptScore W4386938161C187736073 @default.
- W4386938161 hasConceptScore W4386938161C204321447 @default.
- W4386938161 hasConceptScore W4386938161C2780451532 @default.
- W4386938161 hasConceptScore W4386938161C38652104 @default.
- W4386938161 hasConceptScore W4386938161C41008148 @default.
- W4386938161 hasConceptScore W4386938161C41895202 @default.
- W4386938161 hasConceptScore W4386938161C48145219 @default.
- W4386938161 hasConceptScore W4386938161C62520636 @default.
- W4386938161 hasConceptScore W4386938161C66322947 @default.
- W4386938161 hasConceptScore W4386938161C66402592 @default.
- W4386938161 hasConceptScore W4386938161C90805587 @default.
- W4386938161 hasLocation W43869381611 @default.
- W4386938161 hasOpenAccess W4386938161 @default.
- W4386938161 hasPrimaryLocation W43869381611 @default.
- W4386938161 hasRelatedWork W2335882425 @default.
- W4386938161 hasRelatedWork W3029282628 @default.
- W4386938161 hasRelatedWork W3037162118 @default.
- W4386938161 hasRelatedWork W3192794374 @default.
- W4386938161 hasRelatedWork W4221139500 @default.
- W4386938161 hasRelatedWork W4285666859 @default.
- W4386938161 hasRelatedWork W4287775347 @default.
- W4386938161 hasRelatedWork W4362613237 @default.
- W4386938161 hasRelatedWork W4386076661 @default.
- W4386938161 hasRelatedWork W4226206785 @default.