Matches in SemOpenAlex for { <https://semopenalex.org/work/W2912473624> ?p ?o ?g. }
- W2912473624 abstract "It is intuitive that NLP tasks for logographic languages like Chinese should benefit from the use of the glyph information in those languages. However, due to the lack of rich pictographic evidence in glyphs and the weak generalization ability of standard computer vision models on character data, an effective way to utilize the glyph information remains to be found. In this paper, we address this gap by presenting Glyce, the glyph-vectors for Chinese character representations. We make three major innovations: (1) We use historical Chinese scripts (e.g., bronzeware script, seal script, traditional Chinese, etc) to enrich the pictographic evidence in characters; (2) We design CNN structures (called tianzege-CNN) tailored to Chinese character image processing; and (3) We use image-classification as an auxiliary task in a multi-task learning setup to increase the model's ability to generalize. We show that glyph-based models are able to consistently outperform word/char ID-based models in a wide range of Chinese NLP tasks. We are able to set new state-of-the-art results for a variety of Chinese NLP tasks, including tagging (NER, CWS, POS), sentence pair classification, single sentence classification tasks, dependency parsing, and semantic role labeling. For example, the proposed model achieves an F1 score of 80.6 on the OntoNotes dataset of NER, +1.5 over BERT; it achieves an almost perfect accuracy of 99.8% on the Fudan corpus for text classification. Code found at https://github.com/ShannonAI/glyce." @default.
- W2912473624 created "2019-02-21" @default.
- W2912473624 creator A5003511813 @default.
- W2912473624 creator A5014777126 @default.
- W2912473624 creator A5023207081 @default.
- W2912473624 creator A5025603106 @default.
- W2912473624 creator A5053347296 @default.
- W2912473624 creator A5058189738 @default.
- W2912473624 creator A5074612639 @default.
- W2912473624 creator A5077851706 @default.
- W2912473624 creator A5081449019 @default.
- W2912473624 creator A5086529957 @default.
- W2912473624 date "2019-01-29" @default.
- W2912473624 modified "2023-10-17" @default.
- W2912473624 title "Glyce: Glyph-vectors for Chinese Character Representations" @default.
- W2912473624 cites W1575907248 @default.
- W2912473624 cites W1594229598 @default.
- W2912473624 cites W1840435438 @default.
- W2912473624 cites W1938755728 @default.
- W2912473624 cites W2101105183 @default.
- W2912473624 cites W2108598243 @default.
- W2912473624 cites W2166706824 @default.
- W2912473624 cites W2183341477 @default.
- W2912473624 cites W2194775991 @default.
- W2912473624 cites W2250539671 @default.
- W2912473624 cites W2250767751 @default.
- W2912473624 cites W2250861254 @default.
- W2912473624 cites W2251131401 @default.
- W2912473624 cites W2251293245 @default.
- W2912473624 cites W2251500379 @default.
- W2912473624 cites W2251811146 @default.
- W2912473624 cites W2251939518 @default.
- W2912473624 cites W2311921240 @default.
- W2912473624 cites W2339995566 @default.
- W2912473624 cites W2397198482 @default.
- W2912473624 cites W2436788615 @default.
- W2912473624 cites W2468328197 @default.
- W2912473624 cites W2468476969 @default.
- W2912473624 cites W2518668950 @default.
- W2912473624 cites W2552110825 @default.
- W2912473624 cites W2566150155 @default.
- W2912473624 cites W2593833795 @default.
- W2912473624 cites W2608256743 @default.
- W2912473624 cites W2609232568 @default.
- W2912473624 cites W2738591886 @default.
- W2912473624 cites W2740603853 @default.
- W2912473624 cites W2743935014 @default.
- W2912473624 cites W2745470944 @default.
- W2912473624 cites W2752461400 @default.
- W2912473624 cites W2756386045 @default.
- W2912473624 cites W2757350179 @default.
- W2912473624 cites W2759366113 @default.
- W2912473624 cites W2778955544 @default.
- W2912473624 cites W2787560479 @default.
- W2912473624 cites W2788009253 @default.
- W2912473624 cites W2798304389 @default.
- W2912473624 cites W2799090016 @default.
- W2912473624 cites W2801060378 @default.
- W2912473624 cites W2876111955 @default.
- W2912473624 cites W2877808116 @default.
- W2912473624 cites W2887282701 @default.
- W2912473624 cites W2888293024 @default.
- W2912473624 cites W2889968917 @default.
- W2912473624 cites W2896649846 @default.
- W2912473624 cites W2899395607 @default.
- W2912473624 cites W2922018002 @default.
- W2912473624 cites W2949777170 @default.
- W2912473624 cites W2949952998 @default.
- W2912473624 cites W2950133940 @default.
- W2912473624 cites W2950886545 @default.
- W2912473624 cites W2951449747 @default.
- W2912473624 cites W2951545716 @default.
- W2912473624 cites W2951941802 @default.
- W2912473624 cites W2951975363 @default.
- W2912473624 cites W2952230511 @default.
- W2912473624 cites W2952935105 @default.
- W2912473624 cites W2953391617 @default.
- W2912473624 cites W2962902328 @default.
- W2912473624 cites W2962904552 @default.
- W2912473624 cites W2963341956 @default.
- W2912473624 cites W2963355447 @default.
- W2912473624 cites W2963403868 @default.
- W2912473624 cites W2963913268 @default.
- W2912473624 cites W2964093505 @default.
- W2912473624 cites W2964325863 @default.
- W2912473624 doi "https://doi.org/10.48550/arxiv.1901.10125" @default.
- W2912473624 hasPublicationYear "2019" @default.
- W2912473624 type Work @default.
- W2912473624 sameAs 2912473624 @default.
- W2912473624 citedByCount "9" @default.
- W2912473624 countsByYear W29124736242019 @default.
- W2912473624 countsByYear W29124736242021 @default.
- W2912473624 crossrefType "posted-content" @default.
- W2912473624 hasAuthorship W2912473624A5003511813 @default.
- W2912473624 hasAuthorship W2912473624A5014777126 @default.
- W2912473624 hasAuthorship W2912473624A5023207081 @default.
- W2912473624 hasAuthorship W2912473624A5025603106 @default.
- W2912473624 hasAuthorship W2912473624A5053347296 @default.
- W2912473624 hasAuthorship W2912473624A5058189738 @default.
- W2912473624 hasAuthorship W2912473624A5074612639 @default.