Everyone interested in this topic is welcomed to join us. the following is the abstract for this week\u2019s paper.<\/p>\n\n\n\n

\n\t

\n\t\tGlyce: Glyph-vectors for Chinese Character Representations\n\t<\/p>\n\t

\n\t\tYuxian Meng, Wei\nWu, Qinghong Han, Muyu Li, Xiaoya Li, Jie Mei, Ping Nie, Xiaofei Sun and Jiwei\nLi\n\t<\/p>\n\t

\n\t\tAbstract\n\t<\/p>\n\t

\n\t\t It is intuitive that NLP tasks for\nlogographic languages like Chinese should benefit from the use of the glyph\ninformation in those languages. However, due to the lack of rich pictographic\nevidence in glyphs and the weak generalization ability of standard computer\nvision models on character data, an effective way to utilize the glyph\ninformation remains to be found.<\/span> \n\t<\/p>\n\t

\n\t\t In this paper, we address this gap by\npresenting the Glyce, the glyph-vectors for Chinese character representations.\nWe make three major innovations: (1) We use historical Chinese scripts (e.g.,\nbronzeware script, seal script, traditional Chinese, etc) to enrich the\npictographic evidence in characters; (2) We design CNN structures tailored to\nChinese character image processing; and (3) We use image-classification as an\nauxiliary task in a multi-task learning setup to increase the model\u2019s ability\nto generalize.<\/span> \n\t<\/p>\n\t

\n\t\t For the first time, we show that\nglyph-based models are able to consistently outperform word\/char ID-based\nmodels in a wide range of Chinese NLP tasks. Using Glyce, we are able to\nachieve the state-of-the-art performances on 13 (almost all) Chinese NLP tasks,\nincluding (1) character-Level language modeling, (2) word-Level language\nmodeling, (3) Chinese word segmentation, (4) name entity recognition, (5)\npart-of-speech tagging, (6) dependency parsing, (7) semantic role labeling, (8)\nsentence semantic similarity, (9) sentence intention identification, (10)\nChinese-English machine translation, (11) sentiment analysis, (12) document\nclassification and (13) discourse parsing.<\/span> \n\t<\/p>\n\t

\n\t\t<\/span> \n\t<\/p>\n<\/div>\n\n\n\n

Glyce- Glyph-vectors for Chinese Character Representations<\/a>\u4e0b\u8f7d<\/a><\/div>\n\n\n\n\n\n\n\n

NLPIR SEMINAR 21st ISSUE COMPLETED<\/strong><\/h2>\n\n\n\n
Last Monday, Li Shen<\/strong> gave a presentation about the paper, Glyce: Glyph-vectors for Chinese Character Representations<\/strong>, and shared some opinion on it.<\/p>\n\n\n\n
Table 3 reports perplexity for each model, along with the number of parameters. The ppl is short for perplexity.
the novelty of this paper is to make use of Chinese charater glyph information in NLP.<\/p>\n","protected":false},"excerpt":{"rendered":"
NLPIR SEMINAR Y2019#8 INTRO In the new s … \u7ee7\u7eed\u9605\u8bfb →<\/span><\/a><\/p>\n","protected":false},"author":862,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[37,38],"tags":[],"_links":{"self":[{"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/posts\/6774"}],"collection":[{"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/users\/862"}],"replies":[{"embeddable":true,"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/comments?post=6774"}],"version-history":[{"count":3,"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/posts\/6774\/revisions"}],"predecessor-version":[{"id":6828,"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/posts\/6774\/revisions\/6828"}],"wp:attachment":[{"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/media?parent=6774"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/categories?post=6774"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.nlpir.org\/wordpress\/wp-json\/wp\/v2\/tags?post=6774"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}

INTRO <\/h3>\n\n\n\n In the new semester, our Lab, Web Search Mining and Security Lab, plans to hold an academic seminar every Monday, and each time a keynote speaker will share understanding of papers on his\/her related research with you.<\/p>\n\n\n\n

INTRO <\/h3>\n\n\n\n
In the new semester, our Lab, Web Search Mining and Security Lab, plans to hold an academic seminar every Monday, and each time a keynote speaker will share understanding of papers on his\/her related research with you.
<\/p>\n\n\n\n