Matches in SemOpenAlex for { <https://semopenalex.org/work/W3206550936> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W3206550936 abstract "Recently, pre-trained language models have gained dramatic progress on grammatical error correction (GEC) task by fine-tuning on a small amount of annotated data. However, the current approaches ignore two problems. On the one hand, the GEC datasets suffer from annotation errors which may impair the performance of the model. On the other hand, the correction difficulty varies across sentences and the generating difficulty of each token within a sentence is inconsistent as well. Therefore, hard and easy samples in GEC task should be treated differently. To address these issues, we propose an uncertainty-aware self-paced learning framework for GEC task. We leverage Bayesian deep learning to mine and filter noisy samples in the training set. Besides, we design a confidence-based self-paced learning strategy to dynamically adjust the loss weights of hard and easy samples. Specifically, we measure the confidence score of the model on the samples at the token-level and the sentence-level, and schedule the training procedure according to the confidence scores. Extensive experiments demonstrate that the proposed approach surpasses the baseline model by 2.0+ point of (F_{0.5}) scores on several GEC datasets and proves the effectiveness of our approach." @default.
- W3206550936 created "2021-10-25" @default.
- W3206550936 creator A5002654282 @default.
- W3206550936 creator A5014400759 @default.
- W3206550936 creator A5030147054 @default.
- W3206550936 creator A5087037080 @default.
- W3206550936 date "2021-01-01" @default.
- W3206550936 modified "2023-09-25" @default.
- W3206550936 title "Uncertainty-Aware Self-paced Learning for Grammatical Error Correction" @default.
- W3206550936 cites W2296073425 @default.
- W3206550936 cites W2470324779 @default.
- W3206550936 cites W2797885244 @default.
- W3206550936 cites W2798416860 @default.
- W3206550936 cites W2890230387 @default.
- W3206550936 cites W2963351448 @default.
- W3206550936 cites W2963516811 @default.
- W3206550936 cites W2963881719 @default.
- W3206550936 cites W2964082031 @default.
- W3206550936 cites W2964187553 @default.
- W3206550936 cites W2970429618 @default.
- W3206550936 cites W2970521905 @default.
- W3206550936 cites W2970744242 @default.
- W3206550936 cites W2970868759 @default.
- W3206550936 cites W2997636815 @default.
- W3206550936 cites W3034999214 @default.
- W3206550936 cites W3035010485 @default.
- W3206550936 cites W3037162118 @default.
- W3206550936 cites W3096648221 @default.
- W3206550936 cites W3101155369 @default.
- W3206550936 cites W3103818060 @default.
- W3206550936 cites W3104669693 @default.
- W3206550936 cites W3165374601 @default.
- W3206550936 doi "https://doi.org/10.1007/978-3-030-88480-2_59" @default.
- W3206550936 hasPublicationYear "2021" @default.
- W3206550936 type Work @default.
- W3206550936 sameAs 3206550936 @default.
- W3206550936 citedByCount "0" @default.
- W3206550936 crossrefType "book-chapter" @default.
- W3206550936 hasAuthorship W3206550936A5002654282 @default.
- W3206550936 hasAuthorship W3206550936A5014400759 @default.
- W3206550936 hasAuthorship W3206550936A5030147054 @default.
- W3206550936 hasAuthorship W3206550936A5087037080 @default.
- W3206550936 hasConcept C111368507 @default.
- W3206550936 hasConcept C119857082 @default.
- W3206550936 hasConcept C12725497 @default.
- W3206550936 hasConcept C127313418 @default.
- W3206550936 hasConcept C137293760 @default.
- W3206550936 hasConcept C153083717 @default.
- W3206550936 hasConcept C154945302 @default.
- W3206550936 hasConcept C162324750 @default.
- W3206550936 hasConcept C177264268 @default.
- W3206550936 hasConcept C187736073 @default.
- W3206550936 hasConcept C199360897 @default.
- W3206550936 hasConcept C204321447 @default.
- W3206550936 hasConcept C2777530160 @default.
- W3206550936 hasConcept C2780451532 @default.
- W3206550936 hasConcept C38652104 @default.
- W3206550936 hasConcept C41008148 @default.
- W3206550936 hasConcept C48145219 @default.
- W3206550936 hasConcept C97541855 @default.
- W3206550936 hasConceptScore W3206550936C111368507 @default.
- W3206550936 hasConceptScore W3206550936C119857082 @default.
- W3206550936 hasConceptScore W3206550936C12725497 @default.
- W3206550936 hasConceptScore W3206550936C127313418 @default.
- W3206550936 hasConceptScore W3206550936C137293760 @default.
- W3206550936 hasConceptScore W3206550936C153083717 @default.
- W3206550936 hasConceptScore W3206550936C154945302 @default.
- W3206550936 hasConceptScore W3206550936C162324750 @default.
- W3206550936 hasConceptScore W3206550936C177264268 @default.
- W3206550936 hasConceptScore W3206550936C187736073 @default.
- W3206550936 hasConceptScore W3206550936C199360897 @default.
- W3206550936 hasConceptScore W3206550936C204321447 @default.
- W3206550936 hasConceptScore W3206550936C2777530160 @default.
- W3206550936 hasConceptScore W3206550936C2780451532 @default.
- W3206550936 hasConceptScore W3206550936C38652104 @default.
- W3206550936 hasConceptScore W3206550936C41008148 @default.
- W3206550936 hasConceptScore W3206550936C48145219 @default.
- W3206550936 hasConceptScore W3206550936C97541855 @default.
- W3206550936 hasLocation W32065509361 @default.
- W3206550936 hasOpenAccess W3206550936 @default.
- W3206550936 hasPrimaryLocation W32065509361 @default.
- W3206550936 hasRelatedWork W10961682 @default.
- W3206550936 hasRelatedWork W13978995 @default.
- W3206550936 hasRelatedWork W193554 @default.
- W3206550936 hasRelatedWork W4244564 @default.
- W3206550936 hasRelatedWork W4784704 @default.
- W3206550936 hasRelatedWork W6633147 @default.
- W3206550936 hasRelatedWork W881173 @default.
- W3206550936 hasRelatedWork W9218159 @default.
- W3206550936 hasRelatedWork W9840409 @default.
- W3206550936 hasRelatedWork W13002482 @default.
- W3206550936 isParatext "false" @default.
- W3206550936 isRetracted "false" @default.
- W3206550936 magId "3206550936" @default.
- W3206550936 workType "book-chapter" @default.