FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3210811241> ?p ?o ?g. }

• W3210811241 abstract "Though there has been a large body of recent works in language modeling (LM) for high resource languages such as English and Chinese, the area is still unexplored for low resource languages like Bengali and Hindi. We propose an end to end trainable memory efficient CNN architecture named CoCNN to handle specific characteristics such as high inflection, morphological richness, flexible word order and phonetical spelling errors of Bengali and Hindi. In particular, we introduce two learnable convolutional sub-models at word and at sentence level that are end to end trainable. We show that state-of-the-art (SOTA) Transformer models including pretrained BERT do not necessarily yield the best performance for Bengali and Hindi. CoCNN outperforms pretrained BERT with 16X less parameters, and it achieves much better performance than SOTA LSTM models on multiple real-world datasets. This is the first study on the effectiveness of different architectures drawn from three deep learning paradigms - Convolution, Recurrent, and Transformer neural nets for modeling two widely used languages, Bengali and Hindi." @default.
• W3210811241 created "2021-11-08" @default.
• W3210811241 creator A5001619124 @default.
• W3210811241 creator A5019261399 @default.
• W3210811241 creator A5022817303 @default.
• W3210811241 creator A5032790646 @default.
• W3210811241 creator A5033438760 @default.
• W3210811241 creator A5036160043 @default.
• W3210811241 date "2021-10-25" @default.
• W3210811241 modified "2023-09-27" @default.
• W3210811241 title "Revisiting CNN for Highly Inflected Bengali and Hindi Language Modeling" @default.
• W3210811241 cites W1489971870 @default.
• W3210811241 cites W1902237438 @default.
• W3210811241 cites W1904365287 @default.
• W3210811241 cites W2049031033 @default.
• W3210811241 cites W2133564696 @default.
• W3210811241 cites W2171928131 @default.
• W3210811241 cites W2436788615 @default.
• W3210811241 cites W2493916176 @default.
• W3210811241 cites W2539719263 @default.
• W3210811241 cites W2553941291 @default.
• W3210811241 cites W2566563465 @default.
• W3210811241 cites W2598880630 @default.
• W3210811241 cites W2626778328 @default.
• W3210811241 cites W2743945814 @default.
• W3210811241 cites W2798760675 @default.
• W3210811241 cites W2883158411 @default.
• W3210811241 cites W2943845043 @default.
• W3210811241 cites W2951559648 @default.
• W3210811241 cites W2952258850 @default.
• W3210811241 cites W295894637 @default.
• W3210811241 cites W2962695106 @default.
• W3210811241 cites W2963088785 @default.
• W3210811241 cites W2963341956 @default.
• W3210811241 cites W2965373594 @default.
• W3210811241 cites W2970213198 @default.
• W3210811241 cites W2995463029 @default.
• W3210811241 cites W3009651615 @default.
• W3210811241 hasPublicationYear "2021" @default.
• W3210811241 type Work @default.
• W3210811241 sameAs 3210811241 @default.
• W3210811241 citedByCount "0" @default.
• W3210811241 crossrefType "posted-content" @default.
• W3210811241 hasAuthorship W3210811241A5001619124 @default.
• W3210811241 hasAuthorship W3210811241A5019261399 @default.
• W3210811241 hasAuthorship W3210811241A5022817303 @default.
• W3210811241 hasAuthorship W3210811241A5032790646 @default.
• W3210811241 hasAuthorship W3210811241A5033438760 @default.
• W3210811241 hasAuthorship W3210811241A5036160043 @default.
• W3210811241 hasConcept C121332964 @default.
• W3210811241 hasConcept C137293760 @default.
• W3210811241 hasConcept C154945302 @default.
• W3210811241 hasConcept C159403335 @default.
• W3210811241 hasConcept C165801399 @default.
• W3210811241 hasConcept C19235068 @default.
• W3210811241 hasConcept C204321447 @default.
• W3210811241 hasConcept C2777530160 @default.
• W3210811241 hasConcept C28490314 @default.
• W3210811241 hasConcept C41008148 @default.
• W3210811241 hasConcept C519982507 @default.
• W3210811241 hasConcept C62520636 @default.
• W3210811241 hasConcept C66322947 @default.
• W3210811241 hasConceptScore W3210811241C121332964 @default.
• W3210811241 hasConceptScore W3210811241C137293760 @default.
• W3210811241 hasConceptScore W3210811241C154945302 @default.
• W3210811241 hasConceptScore W3210811241C159403335 @default.
• W3210811241 hasConceptScore W3210811241C165801399 @default.
• W3210811241 hasConceptScore W3210811241C19235068 @default.
• W3210811241 hasConceptScore W3210811241C204321447 @default.
• W3210811241 hasConceptScore W3210811241C2777530160 @default.
• W3210811241 hasConceptScore W3210811241C28490314 @default.
• W3210811241 hasConceptScore W3210811241C41008148 @default.
• W3210811241 hasConceptScore W3210811241C519982507 @default.
• W3210811241 hasConceptScore W3210811241C62520636 @default.
• W3210811241 hasConceptScore W3210811241C66322947 @default.
• W3210811241 hasLocation W32108112411 @default.
• W3210811241 hasOpenAccess W3210811241 @default.
• W3210811241 hasPrimaryLocation W32108112411 @default.
• W3210811241 hasRelatedWork W2013997717 @default.
• W3210811241 hasRelatedWork W2060413777 @default.
• W3210811241 hasRelatedWork W2251297821 @default.
• W3210811241 hasRelatedWork W2575131341 @default.
• W3210811241 hasRelatedWork W2741095062 @default.
• W3210811241 hasRelatedWork W2888831408 @default.
• W3210811241 hasRelatedWork W2908700150 @default.
• W3210811241 hasRelatedWork W2963739452 @default.
• W3210811241 hasRelatedWork W2973887018 @default.
• W3210811241 hasRelatedWork W2980609131 @default.
• W3210811241 hasRelatedWork W2997035194 @default.
• W3210811241 hasRelatedWork W3087204350 @default.
• W3210811241 hasRelatedWork W3095271009 @default.
• W3210811241 hasRelatedWork W3120157709 @default.
• W3210811241 hasRelatedWork W3122978153 @default.
• W3210811241 hasRelatedWork W3125631070 @default.
• W3210811241 hasRelatedWork W3199212312 @default.
• W3210811241 hasRelatedWork W3206914982 @default.
• W3210811241 hasRelatedWork W3211544529 @default.
• W3210811241 hasRelatedWork W2511874079 @default.
• W3210811241 isParatext "false" @default.
• W3210811241 isRetracted "false" @default.

• next