Matches in SemOpenAlex for { <https://semopenalex.org/work/W3022112973> ?p ?o ?g. }
- W3022112973 abstract "We compare different models for low resource multi-task sequence tagging that leverage dependencies between label sequences for different tasks. Our analysis is aimed at datasets where each example has labels for multiple tasks. Current approaches use either a separate model for each task or standard multi-task learning to learn shared feature representations. However, these approaches ignore correlations between label sequences, which can provide important information in settings with small training datasets. To analyze which scenarios can profit from modeling dependencies between labels in different tasks, we revisit dynamic conditional random fields (CRFs) and combine them with deep neural networks. We compare single-task, multi-task and dynamic CRF setups for three diverse datasets at both sentence and document levels in English and German low resource scenarios. We show that including silver labels from pretrained part-of-speech taggers as auxiliary tasks can improve performance on downstream tasks. We find that especially in low-resource scenarios, the explicit modeling of inter-dependencies between task predictions outperforms single-task as well as standard multi-task models." @default.
- W3022112973 created "2020-05-13" @default.
- W3022112973 creator A5024983536 @default.
- W3022112973 creator A5027450194 @default.
- W3022112973 creator A5061992028 @default.
- W3022112973 date "2020-05-01" @default.
- W3022112973 modified "2023-09-30" @default.
- W3022112973 title "Low Resource Multi-Task Sequence Tagging - Revisiting Dynamic Conditional Random Fields." @default.
- W3022112973 cites W1498763386 @default.
- W3022112973 cites W1522301498 @default.
- W3022112973 cites W1560512119 @default.
- W3022112973 cites W1682403713 @default.
- W3022112973 cites W1940872118 @default.
- W3022112973 cites W2008784586 @default.
- W3022112973 cites W2038721957 @default.
- W3022112973 cites W2038858953 @default.
- W3022112973 cites W2047477415 @default.
- W3022112973 cites W2060277733 @default.
- W3022112973 cites W2108218871 @default.
- W3022112973 cites W2117130368 @default.
- W3022112973 cites W2129999749 @default.
- W3022112973 cites W2142384583 @default.
- W3022112973 cites W2147880316 @default.
- W3022112973 cites W2152463966 @default.
- W3022112973 cites W2158823144 @default.
- W3022112973 cites W2250539671 @default.
- W3022112973 cites W2296283641 @default.
- W3022112973 cites W2493916176 @default.
- W3022112973 cites W2531207078 @default.
- W3022112973 cites W2592170186 @default.
- W3022112973 cites W2624871570 @default.
- W3022112973 cites W2741608233 @default.
- W3022112973 cites W2742079690 @default.
- W3022112973 cites W2804221886 @default.
- W3022112973 cites W2823875838 @default.
- W3022112973 cites W2880875857 @default.
- W3022112973 cites W2889869899 @default.
- W3022112973 cites W2913340405 @default.
- W3022112973 cites W2923126539 @default.
- W3022112973 cites W2935052563 @default.
- W3022112973 cites W2948086026 @default.
- W3022112973 cites W2950281195 @default.
- W3022112973 cites W2962676330 @default.
- W3022112973 cites W2962739339 @default.
- W3022112973 cites W2962897020 @default.
- W3022112973 cites W2963153906 @default.
- W3022112973 cites W2963168538 @default.
- W3022112973 cites W2963211188 @default.
- W3022112973 cites W2963341956 @default.
- W3022112973 cites W2963403868 @default.
- W3022112973 cites W2963563735 @default.
- W3022112973 cites W2963940534 @default.
- W3022112973 cites W2964298396 @default.
- W3022112973 cites W2964303773 @default.
- W3022112973 cites W2981573048 @default.
- W3022112973 cites W2997792775 @default.
- W3022112973 cites W3153675281 @default.
- W3022112973 cites W630883834 @default.
- W3022112973 hasPublicationYear "2020" @default.
- W3022112973 type Work @default.
- W3022112973 sameAs 3022112973 @default.
- W3022112973 citedByCount "3" @default.
- W3022112973 countsByYear W30221129732020 @default.
- W3022112973 countsByYear W30221129732021 @default.
- W3022112973 crossrefType "posted-content" @default.
- W3022112973 hasAuthorship W3022112973A5024983536 @default.
- W3022112973 hasAuthorship W3022112973A5027450194 @default.
- W3022112973 hasAuthorship W3022112973A5061992028 @default.
- W3022112973 hasConcept C119857082 @default.
- W3022112973 hasConcept C152565575 @default.
- W3022112973 hasConcept C153083717 @default.
- W3022112973 hasConcept C154945302 @default.
- W3022112973 hasConcept C162324750 @default.
- W3022112973 hasConcept C187736073 @default.
- W3022112973 hasConcept C204321447 @default.
- W3022112973 hasConcept C2775953691 @default.
- W3022112973 hasConcept C2777530160 @default.
- W3022112973 hasConcept C2780451532 @default.
- W3022112973 hasConcept C35639132 @default.
- W3022112973 hasConcept C41008148 @default.
- W3022112973 hasConcept C95623464 @default.
- W3022112973 hasConceptScore W3022112973C119857082 @default.
- W3022112973 hasConceptScore W3022112973C152565575 @default.
- W3022112973 hasConceptScore W3022112973C153083717 @default.
- W3022112973 hasConceptScore W3022112973C154945302 @default.
- W3022112973 hasConceptScore W3022112973C162324750 @default.
- W3022112973 hasConceptScore W3022112973C187736073 @default.
- W3022112973 hasConceptScore W3022112973C204321447 @default.
- W3022112973 hasConceptScore W3022112973C2775953691 @default.
- W3022112973 hasConceptScore W3022112973C2777530160 @default.
- W3022112973 hasConceptScore W3022112973C2780451532 @default.
- W3022112973 hasConceptScore W3022112973C35639132 @default.
- W3022112973 hasConceptScore W3022112973C41008148 @default.
- W3022112973 hasConceptScore W3022112973C95623464 @default.
- W3022112973 hasLocation W30221129731 @default.
- W3022112973 hasOpenAccess W3022112973 @default.
- W3022112973 hasPrimaryLocation W30221129731 @default.
- W3022112973 hasRelatedWork W2508857598 @default.
- W3022112973 hasRelatedWork W2765742186 @default.
- W3022112973 hasRelatedWork W2786157125 @default.