Matches in SemOpenAlex for { <https://semopenalex.org/work/W3120689745> ?p ?o ?g. }
Showing items 1 to 99 of
99
with 100 items per page.
- W3120689745 abstract "Extreme Multi-label text Classification (XMC) is a task of finding the most relevant labels from a large label set. Nowadays deep learning-based methods have shown significant success in XMC. However, the existing methods (e.g., AttentionXML and X-Transformer etc) still suffer from 1) combining several models to train and predict for one dataset, and 2) sampling negative labels statically during the process of training label ranking model, which reduces both the efficiency and accuracy of the model. To address the above problems, we proposed LightXML, which adopts end-to-end training and dynamic negative labels sampling. In LightXML, we use generative cooperative networks to recall and rank labels, in which label recalling part generates negative and positive labels, and label ranking part distinguishes positive labels from these labels. Through these networks, negative labels are sampled dynamically during label ranking part training by feeding with the same text representation. Extensive experiments show that LightXML outperforms state-of-the-art methods in five extreme multi-label datasets with much smaller model size and lower computational complexity. In particular, on the Amazon dataset with 670K labels, LightXML can reduce the model size up to 72% compared to AttentionXML." @default.
- W3120689745 created "2021-01-18" @default.
- W3120689745 creator A5006320644 @default.
- W3120689745 creator A5041921496 @default.
- W3120689745 creator A5057010946 @default.
- W3120689745 creator A5081275566 @default.
- W3120689745 creator A5081856566 @default.
- W3120689745 creator A5090606731 @default.
- W3120689745 date "2021-01-09" @default.
- W3120689745 modified "2023-09-27" @default.
- W3120689745 title "LightXML: Transformer with Dynamic Negative Sampling for High-Performance Extreme Multi-label Text Classification" @default.
- W3120689745 cites W1522301498 @default.
- W3120689745 cites W1781770377 @default.
- W3120689745 cites W2061873838 @default.
- W3120689745 cites W2068074736 @default.
- W3120689745 cites W2100743709 @default.
- W3120689745 cites W2183087644 @default.
- W3120689745 cites W21938781 @default.
- W3120689745 cites W2461743311 @default.
- W3120689745 cites W2520348554 @default.
- W3120689745 cites W2739996966 @default.
- W3120689745 cites W2743021690 @default.
- W3120689745 cites W2744136723 @default.
- W3120689745 cites W2788125153 @default.
- W3120689745 cites W2792287754 @default.
- W3120689745 cites W2890634844 @default.
- W3120689745 cites W2921113176 @default.
- W3120689745 cites W2936191805 @default.
- W3120689745 cites W2963341956 @default.
- W3120689745 cites W2963403868 @default.
- W3120689745 cites W2965373594 @default.
- W3120689745 cites W2970574105 @default.
- W3120689745 cites W2970597249 @default.
- W3120689745 cites W3037422790 @default.
- W3120689745 doi "https://doi.org/10.48550/arxiv.2101.03305" @default.
- W3120689745 hasPublicationYear "2021" @default.
- W3120689745 type Work @default.
- W3120689745 sameAs 3120689745 @default.
- W3120689745 citedByCount "6" @default.
- W3120689745 countsByYear W31206897452021 @default.
- W3120689745 crossrefType "posted-content" @default.
- W3120689745 hasAuthorship W3120689745A5006320644 @default.
- W3120689745 hasAuthorship W3120689745A5041921496 @default.
- W3120689745 hasAuthorship W3120689745A5057010946 @default.
- W3120689745 hasAuthorship W3120689745A5081275566 @default.
- W3120689745 hasAuthorship W3120689745A5081856566 @default.
- W3120689745 hasAuthorship W3120689745A5090606731 @default.
- W3120689745 hasBestOaLocation W31206897451 @default.
- W3120689745 hasConcept C100660578 @default.
- W3120689745 hasConcept C119599485 @default.
- W3120689745 hasConcept C119857082 @default.
- W3120689745 hasConcept C124101348 @default.
- W3120689745 hasConcept C127413603 @default.
- W3120689745 hasConcept C138885662 @default.
- W3120689745 hasConcept C153180895 @default.
- W3120689745 hasConcept C154945302 @default.
- W3120689745 hasConcept C165801399 @default.
- W3120689745 hasConcept C177264268 @default.
- W3120689745 hasConcept C189430467 @default.
- W3120689745 hasConcept C199360897 @default.
- W3120689745 hasConcept C201995342 @default.
- W3120689745 hasConcept C2780451532 @default.
- W3120689745 hasConcept C41008148 @default.
- W3120689745 hasConcept C41895202 @default.
- W3120689745 hasConcept C66322947 @default.
- W3120689745 hasConceptScore W3120689745C100660578 @default.
- W3120689745 hasConceptScore W3120689745C119599485 @default.
- W3120689745 hasConceptScore W3120689745C119857082 @default.
- W3120689745 hasConceptScore W3120689745C124101348 @default.
- W3120689745 hasConceptScore W3120689745C127413603 @default.
- W3120689745 hasConceptScore W3120689745C138885662 @default.
- W3120689745 hasConceptScore W3120689745C153180895 @default.
- W3120689745 hasConceptScore W3120689745C154945302 @default.
- W3120689745 hasConceptScore W3120689745C165801399 @default.
- W3120689745 hasConceptScore W3120689745C177264268 @default.
- W3120689745 hasConceptScore W3120689745C189430467 @default.
- W3120689745 hasConceptScore W3120689745C199360897 @default.
- W3120689745 hasConceptScore W3120689745C201995342 @default.
- W3120689745 hasConceptScore W3120689745C2780451532 @default.
- W3120689745 hasConceptScore W3120689745C41008148 @default.
- W3120689745 hasConceptScore W3120689745C41895202 @default.
- W3120689745 hasConceptScore W3120689745C66322947 @default.
- W3120689745 hasLocation W31206897451 @default.
- W3120689745 hasOpenAccess W3120689745 @default.
- W3120689745 hasPrimaryLocation W31206897451 @default.
- W3120689745 hasRelatedWork W2033914206 @default.
- W3120689745 hasRelatedWork W2138279922 @default.
- W3120689745 hasRelatedWork W2961085424 @default.
- W3120689745 hasRelatedWork W3046775127 @default.
- W3120689745 hasRelatedWork W3107474891 @default.
- W3120689745 hasRelatedWork W3136151093 @default.
- W3120689745 hasRelatedWork W4205958290 @default.
- W3120689745 hasRelatedWork W4283077296 @default.
- W3120689745 hasRelatedWork W4286629047 @default.
- W3120689745 hasRelatedWork W4224009465 @default.
- W3120689745 isParatext "false" @default.
- W3120689745 isRetracted "false" @default.
- W3120689745 magId "3120689745" @default.
- W3120689745 workType "article" @default.