FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2940743083> ?p ?o ?g. }

• W2940743083 endingPage "46582" @default.
• W2940743083 startingPage "46575" @default.
• W2940743083 abstract "Sequence labeling models with recurrent neural network variants, such as long short-term memory (LSTM) and gated recurrent unit (GRU), show promising performance on several natural language processing (NLP) problems, including named-entity recognition (NER). Most existing models utilize word embeddings for capturing similarities between words. However, they lag when handling previously unobserved or infrequent words. Moreover, the attention mechanism has been used to improve sequence labeling tasks. In this paper, we propose an efficient multi-attention layer system for the Arabic named-entity recognition (ANER) task. In addition to word-level embeddings, we adopt character-level embeddings and combine them via an embedding-level attention mechanism. The output is fed into an encoder unit with bidirectional-LSTM, followed by another self-attention layer that is used to boost the system performance. Our model achieves approximately matched F1 score of 91% on the “ANERCorpus.” The overall experimental results demonstrate that our method is superior to other systems. Our approach using multi-layer attention mechanism yields a new state-of-the-art result for the ANER." @default.
• W2940743083 created "2019-05-03" @default.
• W2940743083 creator A5025944532 @default.
• W2940743083 creator A5054496227 @default.
• W2940743083 creator A5072156142 @default.
• W2940743083 date "2019-01-01" @default.
• W2940743083 modified "2023-10-02" @default.
• W2940743083 title "Boosting Arabic Named-Entity Recognition With Multi-Attention Layer" @default.
• W2940743083 cites W1580467103 @default.
• W2940743083 cites W1728883788 @default.
• W2940743083 cites W1860935423 @default.
• W2940743083 cites W1899794420 @default.
• W2940743083 cites W1902237438 @default.
• W2940743083 cites W1984708705 @default.
• W2940743083 cites W1989420656 @default.
• W2940743083 cites W2020278455 @default.
• W2940743083 cites W2033599040 @default.
• W2940743083 cites W2116489716 @default.
• W2940743083 cites W2123671373 @default.
• W2940743083 cites W2138780451 @default.
• W2940743083 cites W2250598139 @default.
• W2940743083 cites W2296283641 @default.
• W2940743083 cites W2469314752 @default.
• W2940743083 cites W2488984245 @default.
• W2940743083 cites W2590462354 @default.
• W2940743083 cites W2600659824 @default.
• W2940743083 cites W2734608416 @default.
• W2940743083 cites W2761525601 @default.
• W2940743083 cites W2767784948 @default.
• W2940743083 cites W2773312050 @default.
• W2940743083 cites W2777949341 @default.
• W2940743083 cites W2797817943 @default.
• W2940743083 cites W2800965800 @default.
• W2940743083 cites W2900788028 @default.
• W2940743083 cites W2963140597 @default.
• W2940743083 cites W2963625095 @default.
• W2940743083 cites W2964026782 @default.
• W2940743083 cites W4237530236 @default.
• W2940743083 cites W2174372777 @default.
• W2940743083 doi "https://doi.org/10.1109/access.2019.2909641" @default.
• W2940743083 hasPublicationYear "2019" @default.
• W2940743083 type Work @default.
• W2940743083 sameAs 2940743083 @default.
• W2940743083 citedByCount "15" @default.
• W2940743083 countsByYear W29407430832019 @default.
• W2940743083 countsByYear W29407430832020 @default.
• W2940743083 countsByYear W29407430832021 @default.
• W2940743083 countsByYear W29407430832022 @default.
• W2940743083 countsByYear W29407430832023 @default.
• W2940743083 crossrefType "journal-article" @default.
• W2940743083 hasAuthorship W2940743083A5025944532 @default.
• W2940743083 hasAuthorship W2940743083A5054496227 @default.
• W2940743083 hasAuthorship W2940743083A5072156142 @default.
• W2940743083 hasBestOaLocation W29407430831 @default.
• W2940743083 hasConcept C111919701 @default.
• W2940743083 hasConcept C118505674 @default.
• W2940743083 hasConcept C138885662 @default.
• W2940743083 hasConcept C147168706 @default.
• W2940743083 hasConcept C154945302 @default.
• W2940743083 hasConcept C162324750 @default.
• W2940743083 hasConcept C178790620 @default.
• W2940743083 hasConcept C185592680 @default.
• W2940743083 hasConcept C187736073 @default.
• W2940743083 hasConcept C204321447 @default.
• W2940743083 hasConcept C2524010 @default.
• W2940743083 hasConcept C2777462759 @default.
• W2940743083 hasConcept C2778112365 @default.
• W2940743083 hasConcept C2779135771 @default.
• W2940743083 hasConcept C2779227376 @default.
• W2940743083 hasConcept C2780451532 @default.
• W2940743083 hasConcept C2780861071 @default.
• W2940743083 hasConcept C28490314 @default.
• W2940743083 hasConcept C33923547 @default.
• W2940743083 hasConcept C35639132 @default.
• W2940743083 hasConcept C41008148 @default.
• W2940743083 hasConcept C41608201 @default.
• W2940743083 hasConcept C41895202 @default.
• W2940743083 hasConcept C46686674 @default.
• W2940743083 hasConcept C50644808 @default.
• W2940743083 hasConcept C54355233 @default.
• W2940743083 hasConcept C86803240 @default.
• W2940743083 hasConcept C90805587 @default.
• W2940743083 hasConceptScore W2940743083C111919701 @default.
• W2940743083 hasConceptScore W2940743083C118505674 @default.
• W2940743083 hasConceptScore W2940743083C138885662 @default.
• W2940743083 hasConceptScore W2940743083C147168706 @default.
• W2940743083 hasConceptScore W2940743083C154945302 @default.
• W2940743083 hasConceptScore W2940743083C162324750 @default.
• W2940743083 hasConceptScore W2940743083C178790620 @default.
• W2940743083 hasConceptScore W2940743083C185592680 @default.
• W2940743083 hasConceptScore W2940743083C187736073 @default.
• W2940743083 hasConceptScore W2940743083C204321447 @default.
• W2940743083 hasConceptScore W2940743083C2524010 @default.
• W2940743083 hasConceptScore W2940743083C2777462759 @default.
• W2940743083 hasConceptScore W2940743083C2778112365 @default.
• W2940743083 hasConceptScore W2940743083C2779135771 @default.
• W2940743083 hasConceptScore W2940743083C2779227376 @default.
• W2940743083 hasConceptScore W2940743083C2780451532 @default.

• next