[B! NLP] stibbarã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

stibbar id:stibbar

NLPã«é–¢ã™ã‚‹stibbarã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (29)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

æ„Ÿæƒ…åˆ†æžã¨ã¯? - æ„Ÿæƒ…åˆ†æžã®èª¬æ˜Ž - AWS
æ„Ÿæƒ…åˆ†æžã¯ã€ã‚ªãƒ”ãƒ‹ã‚ªãƒ³ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã¨ã‚‚å‘¼ã°ã‚Œã€ä¼æ¥ãŒè£½å“ã‚„ã‚µãƒ¼ãƒ“ã‚¹ã‚’æ”¹å–„ã™ã‚‹ã®ã«å½¹ç«‹ã¤é‡è¦ãªãƒ“ã‚¸ãƒã‚¹ã‚¤ãƒ³ãƒ†ãƒªã‚¸ã‚§ãƒ³ã‚¹ãƒ„ãƒ¼ãƒ«ã§ã™ã€‚æ„Ÿæƒ…åˆ†æžã®ãƒ¡ãƒªãƒƒãƒˆã‚’ã„ãã¤ã‹ä»¥ä¸‹ã«ç¤ºã—ã¾ã™ã€‚ å®¢è¦³çš„ãªã‚¤ãƒ³ã‚µã‚¤ãƒˆã‚’æä¾›ã™ã‚‹ ä¼æ¥ã¯ã€äººå·¥çŸ¥èƒ½ (AI) ãƒ™ãƒ¼ã‚¹ã®æ„Ÿæƒ…åˆ†æžãƒ„ãƒ¼ãƒ«ã‚’ä½¿ç”¨ã™ã‚‹ã“ã¨ã§ã€äººé–“ã®ãƒ¬ãƒ“ãƒ¥ãƒ¼æ‹…å½“è€…ã«é–¢é€£ã—ã¦ç”Ÿã˜ã‚‹å€‹äººçš„ãªãƒã‚¤ã‚¢ã‚¹ã‚’å›žé¿ã§ãã¾ã™ã€‚ãã®çµæžœã€ä¼æ¥ã¯é¡§å®¢ã®æ„è¦‹ã‚’åˆ†æžã™ã‚‹éš›ã«ã€ä¸€è²«ã—ãŸå®¢è¦³çš„ãªçµæžœã‚’å¾—ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ ä¾‹ãˆã°ã€æ¬¡ã®æ–‡ã‚’è€ƒãˆã¦ã¿ã¾ã—ã‚‡ã†ã€‚ ãƒ—ãƒã‚»ãƒƒã‚µã®é€Ÿåº¦ã«ã¯é©šã„ã¦ã„ã¾ã™ãŒã€ã™ãã«ç†±ããªã‚‹ã®ã¯æ®‹å¿µã§ã™ã€‚ ãƒžãƒ¼ã‚±ãƒ†ã‚£ãƒ³ã‚°æ‹…å½“è€…ã¯ã€ãƒ¬ãƒ“ãƒ¥ãƒ¼ã«ãŠã„ã¦ä¸æº€ãŒã‚ã‚‹éƒ¨åˆ†ã«ç›®ã‚’ã¤ã¶ã‚Šã€ãƒ—ãƒã‚»ãƒƒã‚µã®ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚’åã£ãŸã‹ãŸã¡ã§è‚¯å®šçš„ã«è©•ä¾¡ã™ã‚‹å¯èƒ½æ€§ãŒã‚ã‚Šã¾ã™ã€‚ã—ã‹ã—ã€æ£ç¢ºãªæ„Ÿæƒ…åˆ†æžãƒ„ãƒ¼ãƒ«ã¯ã€æ„Ÿæƒ…ã‚’å®¢è¦³çš„ã«æŠŠæ¡ã™ã‚‹ãŸã‚ã«ãƒ†ã‚ã‚¹ãƒˆã‚’ã‚½ãƒ¼ãƒˆãŠã‚ˆã³åˆ†é¡žã—ã¾ã™ã€‚ ã‚ˆã‚Šå„ªã‚ŒãŸè£½å“ã¨ã‚µãƒ¼ãƒ“ã‚¹ã‚’æ§‹ç¯‰ã™ã‚‹
stibbar 2023/09/21
https://xp-cloud.jp/blog/2021/09/06/12183/

aws

nlp

comprehend
ãƒªãƒ³ã‚¯
ã€2019å¹´ç‰ˆã€‘è‡ªç„¶è¨€èªžå‡¦ç†ã®ä»£è¡¨çš„ãªãƒ¢ãƒ‡ãƒ«ãƒ»ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ æ™‚ç³»åˆ—ã¾ã¨ã‚ - Qiita
Deleted articles cannot be recovered. Draft of this article would be also deleted. Are you sure you want to delete this article? ã¯ã˜ã‚ã« ã“ã‚Œã¾ã§ã«ææ¡ˆã•ã‚Œã¦ããŸNLPãƒ¢ãƒ‡ãƒ«ä¸€ã¤ä¸€ã¤ã«ã¤ã„ã¦è©³ã—ãçŸ¥ã‚ŠãŸã‹ã£ãŸï¼†æ”¹è‰¯ã®æ´å²ã‚’çŸ¥ã‚ŠãŸã‹ã£ãŸãŸã‚ã€ç‰¹ã«æœ‰åã ã¨æ€ã‚ã‚Œã‚‹ã‚‚ã®ã‚’ã¾ã¨ã‚ã¾ã—ãŸã€‚Mikolovã‚‰ã«ã‚ˆã‚‹RNNLMä»¥é™ã‚’å¯¾è±¡ã¨ã—ã¦ã„ã¾ã™ã€‚ æ³¨ï¼‰ææ¡ˆå¹´ã«ã¤ã„ã¦ã¯ä¸»ã«arXivã®Submission historyã‚’å‚ç…§ã—ã¦ã„ã¾ã™ã€‚è©³ç´°ãªæ—¥ä»˜ã¾ã§åˆ†ã‹ã‚‰ãªã‹ã£ãŸã‚‚ã®ã‚‚ã‚ã‚Šã¾ã™ã€‚ ä»£è¡¨çš„ãªNLPãƒ¢ãƒ‡ãƒ«ãƒ»ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ä¸€è¦§ RNNLM ææ¡ˆå¹´ï¼š2010 ææ¡ˆè€…ï¼šTomÃ¡Å¡ Mikolov1,2, Martin KarafiÃ¡t1, LukÃ¡Å¡ Burget1, Jan â€œHon
stibbar 2022/08/21
qiita

NLP

machine-learning
ãƒªãƒ³ã‚¯
QiitaAPIã¨æ—¥æœ¬èªžå½¢æ…‹ç´ è§£æžAPIã‚’ä½¿ã£ã¦Qiitaã§äººæ°—ã®ãƒ¯ãƒ¼ãƒ‰ã‚’æŽ¢ãã†ï¼
ã“ã‚“ã«ã¡ã¯ã€GMOã‚¢ãƒ‰ãƒžãƒ¼ã‚±ãƒ†ã‚£ãƒ³ã‚° TAXELé–‹ç™ºéƒ¨ã®y-otakiã§ã™ã€‚ ä»Šå›žã¯ã€QiitaAPI + Yahooã®æ—¥æœ¬èªžå½¢æ…‹ç´ è§£æžAPI ã‚’ç”¨ã„ã¦ã€Qiitaã§æµè¡Œã‚Šã®ãƒ¯ãƒ¼ãƒ‰ã‚’æŠ½å‡ºã—ã¦ã¿ã‚ˆã†ã¨æ€ã„ã¾ã™ï¼ ã“ã®è¨˜äº‹ã§ã‚„ã‚‹ã“ã¨ QiitaAPI + Yahooã®æ—¥æœ¬èªžå½¢æ…‹ç´ è§£æžAPI ã‚’ç”¨ã„ã¦ã€Qiitaã®è¨˜äº‹ã‚¿ã‚¤ãƒˆãƒ«ã®å½¢æ…‹ç´ è§£æžã‚’è¡Œã†ã€‚ GASä¸Šã§APIã‚’å‹•ã‹ã—ã€ãã®çµæžœã‚’ã‚·ãƒ¼ãƒˆã«å‡ºåŠ›ã™ã‚‹ã€‚ çµæžœã®è§£æžã‚’è¡Œã„ã€å½¢æ…‹ç´ è§£æžã«ã‚ˆã£ã¦è¨˜äº‹ã‚¿ã‚¤ãƒˆãƒ«ã«å«ã¾ã‚Œã‚„ã™ã„ãƒ¯ãƒ¼ãƒ‰ã‚’æŠ½å‡ºã—ã€è€ƒå¯Ÿã™ã‚‹ã€‚ QiitaAPIã§è¨˜äº‹åã‚’å–å¾—ã—ã‚ˆã†ï¼ ã‚¢ã‚¯ã‚»ã‚¹ãƒˆãƒ¼ã‚¯ãƒ³ã‚’ç™ºè¡Œã—ã¾ã—ã‚‡ã† ã¾ãšã¯ã‚¢ã‚¯ã‚»ã‚¹ãƒˆãƒ¼ã‚¯ãƒ³ã®ç™ºè¡Œã‚’è¡Œã„ã¾ã™ã€‚ ä»Šå›žã¯è¨˜äº‹åã•ãˆå–å¾—ã™ã‚Œã°å•é¡Œãªã„ã®ã§ã€read_qiitaã®ã‚¹ã‚³ãƒ¼ãƒ—ã ã‘æœ‰åŠ¹ã«ã—ã¦ãŠãã¾ã—ã‚‡ã†ã€‚ ã¾ãŸã€ã‚¢ã‚¯ã‚»ã‚¹ãƒˆãƒ¼ã‚¯ãƒ³ã¯ç™ºè¡Œå¾Œã«ç”»é¢ã‚’åˆ‡ã‚Šæ›¿ãˆã‚‹ã¨å†è¡¨ç¤ºå‡ºæ¥ãªããªã‚‹ã®ã§æ³¨æ„ãŒå¿…è¦ã§ã™ã€‚ AP
stibbar 2022/08/14
https://developer.yahoo.co.jp/webapi/jlp/ma/v1/parse.html

NLP
ãƒªãƒ³ã‚¯
å½¢æ…‹ç´ è§£æžã‹ã‚‰æ„Ÿæƒ…åˆ†æžã¾ã§ã€‚æ—¥æœ¬èªžã‚’åˆ†æž/è§£æžã™ã‚‹APIã¾ã¨ã‚ | NTT Communications Developer Portal
æ—¥æœ¬èªžåˆ†æžãªã©ã®ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚’ä½œæˆã™ã‚‹å ´åˆã€å¤§é‡ã®ãƒ‡ãƒ¼ã‚¿ã‚’å…ƒã«å½¢æ…‹ç´ è§£æžãªã©åˆ©ç”¨ã—ã¦åˆ¶ä½œã™ã‚‹ã®ãŒä¸»ä½“ã¨ãªã‚Šã¾ã™ã€‚ã—ã‹ã—ãã®å½¢æ…‹ç´ è§£æžã«ã¤ã„ã¦ã‚‚æ—¢å˜ã®è§£æžç”¨APIã‚’åˆ©ç”¨ã™ã‚‹ã“ã¨ã§ã€è»Šè¼ªã®å†é–‹ç™ºã‚’è¡Œã‚ãªã„ã§æ¸ˆã‚€å¯èƒ½æ€§ãŒã‚ã‚Šã¾ã™ã€‚ä»Šå›žã¯æ—¥æœ¬èªžã®åˆ†æžç³»ã®APIã‚’ã¾ã¨ã‚ã¦ã¿ã¾ã—ãŸã€‚ Yahooã®ãƒ†ã‚ã‚¹ãƒˆè§£æž ãƒ†ã‚ã‚¹ãƒˆè§£æžã§ã¯å¤ãã‹ã‚‰ã‚ã‚‹APIã‚µãƒ¼ãƒ“ã‚¹ã®ä¸€ã¤ã§ã™ã€‚ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³IDæ¯Žã«ãƒªã‚¯ã‚¨ã‚¹ãƒˆæ•°ãªã©ãŒè¦å®šã•ã‚Œã¦ã„ã¾ã™ã®ã§ã€åˆ©ç”¨ã«ã‚ãŸã£ã¦ã¯æ³¨æ„ãŒå¿…è¦ã§ã™ã€‚ æ—¥æœ¬èªžå½¢æ…‹ç´ è§£æž æ—¥æœ¬èªžãƒ†ã‚ã‚¹ãƒˆã‚’å½¢æ…‹ç´ è§£æžã™ã‚‹APIã§ã™ã€‚ç‰¹å¾´çš„ãªã®ã¯å‡ºç¾é »åº¦æƒ…å ±ã‚‚å¾—ã‚‰ã‚Œã‚‹ã“ã¨ã¨ã€å“è©žã‚„åè©žã®å˜ä½ã§ãƒ•ã‚£ãƒ«ã‚¿ã§ãã‚‹ç‚¹ã§ã—ã‚‡ã†ã€‚APIã®å½¢æ…‹ç´ è§£æžã¨ã—ã¦ã¯ä½¿ã„ã‚„ã™ã„ã‚‚ã®ã®ä¸€ã¤ã§ã™ã€‚ ã‹ãªæ¼¢å—å¤‰æ› ãƒãƒ¼ãƒžå—ã€ã²ã‚‰ãŒãªã®ãƒ†ã‚ã‚¹ãƒˆã‹ã‚‰æ–‡ç¯€æ¯Žã«å¤‰æ›å€™è£œã‚’è¿”å´ã™ã‚‹APIã§ã™ã€‚Webä¸Šã‹ã‚‰å…¥åŠ›ã•ã‚ŒãŸæ–‡å—ã‚’æ¼¢å—å¤‰æ›ã™ã‚‹ãªã©ã®ã‚·ãƒ¼ãƒ³ã§åˆ©ç”¨ã§
stibbar 2022/08/14
NLP

Morphological-analysis
ãƒªãƒ³ã‚¯
å½¢æ…‹è«–ã¨ä¾å˜é–¢ä¿‚ãƒ„ãƒªãƒ¼ Â |Â Cloud Natural Language API Â |Â Google Cloud
analyzeSyntax ãƒ¡ã‚½ãƒƒãƒ‰ã¯ã€æŒ‡å®šã•ã‚ŒãŸãƒ†ã‚ã‚¹ãƒˆã®è¨€èªžæ§‹é€ ã«é–¢ã™ã‚‹è©³ç´°ã‚’è¿”ã—ã¾ã™ã€‚ãƒ†ã‚ã‚¹ãƒˆå†…ã®å„ãƒˆãƒ¼ã‚¯ãƒ³ã«ã¤ã„ã¦ã€Natural Language API ã«ã‚ˆã‚Šã€ãã®å†…éƒ¨æ§‹é€ ï¼ˆå½¢æ…‹è«–ï¼‰ã¨æ–‡ã«ãŠã‘ã‚‹ãã®å½¹å‰²ï¼ˆæ§‹æ–‡ï¼‰ã«é–¢ã™ã‚‹æƒ…å ±ãŒæä¾›ã•ã‚Œã¾ã™ã€‚ å½¢æ…‹è«–ã¨ã¯ã€å˜èªžã®å†…éƒ¨æ§‹é€ ã«ã¤ã„ã¦ç ”ç©¶ã—ãŸã‚‚ã®ã§ã™ã€‚å½¢æ…‹è«–ã¯ã€1 ã¤ã®å˜èªžã®æ§‹æˆè¦ç´ ï¼ˆèªžå¹¹ã€èªžç´ ã€æŽ¥é è¾žã€æŽ¥å°¾è¾žãªã©ï¼‰ãŒã©ã®ã‚ˆã†ã«é…ç½®ã¾ãŸã¯å¤‰æ›´ã•ã‚Œã¦ã€ã•ã¾ã–ã¾ãªæ„å‘³ã‚’æŒã¤ã‚ˆã†ã«ãªã‚‹ã‹ã«æ³¨ç›®ã—ã¦ã„ã¾ã™ã€‚ãŸã¨ãˆã°ã€è‹±èªžã§ã¯ã€å¯ç®—åè©žã‚’è¤‡æ•°å½¢ã«ã™ã‚‹å ´åˆã«ã¯ãã®æœ«å°¾ã«ã€Œ-sã€ã‚„ã€Œ-esã€ã‚’ä»˜åŠ ã—ã€å‹•è©žã‚’éŽåŽ»å½¢ã«ã™ã‚‹å ´åˆã«ã¯ãã®æœ«å°¾ã«ã€Œ-dã€ã‚„ã€Œ-edã€ã‚’ä»˜åŠ ã—ã¾ã™ã€‚æŽ¥å°¾è¾žã€Œ-lyã€ã¯ã€å‰¯è©žã‚’ä½œæˆã™ã‚‹ãŸã‚ã«å½¢å®¹è©žã«è¿½åŠ ã•ã‚Œã¾ã™ï¼ˆãŸã¨ãˆã°ã€å½¢å®¹è¯ã€Œhappyã€ã«ã€Œ-lyã€ã‚’è¿½åŠ ã—ã¦ã€å‰¯è©žã€Œhappyfullyã€ã‚’ä½œæˆã™ã‚‹ï¼‰ã€‚ Natural Lang
stibbar 2022/08/14
https://techblog.zozo.com/entry/ios-localization

GCP

NLP
ãƒªãƒ³ã‚¯
è¨€èªžå‡¦ç†100æœ¬ãƒŽãƒƒã‚¯ 2015
è¨€èªžå‡¦ç†100æœ¬ãƒŽãƒƒã‚¯ã¯ï¼Œå®Ÿè·µçš„ãªèª²é¡Œã«å–ã‚Šçµ„ã¿ãªãŒã‚‰ï¼Œãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ï¼Œãƒ‡ãƒ¼ã‚¿åˆ†æžï¼Œç ”ç©¶ã®ã‚¹ã‚ãƒ«ã‚’æ¥½ã—ãç¿’å¾—ã™ã‚‹ã“ã¨ã‚’ç›®æŒ‡ã—ãŸå•é¡Œé›†ã§ã™ å®Ÿç”¨çš„ã§ãƒ¯ã‚¯ãƒ¯ã‚¯ã™ã‚‹ã‚ˆã†ãªé¡Œæã‚’åŽ³é¸ã—ã¾ã—ãŸ è¨€èªžå‡¦ç†ã«åŠ ãˆã¦ï¼Œçµ±è¨ˆã‚„æ©Ÿæ¢°å¦ç¿’ãªã©ã®å‘¨è¾ºåˆ†é‡Žã«ã‚‚è¦ªã—ã‚ã¾ã™ ç ”ç©¶ã‚„ãƒ‡ãƒ¼ã‚¿åˆ†æžã®é€²ã‚æ–¹ï¼Œä½œæ³•ï¼Œã‚¹ã‚ãƒ«ã‚’ä¿®å¾—ã§ãã¾ã™ å•é¡Œã‚’è§£ãã®ã«å¿…è¦ãªãƒ‡ãƒ¼ã‚¿ãƒ»ã‚³ãƒ¼ãƒ‘ã‚¹ã‚’é…å¸ƒã—ã¦ã„ã¾ã™ è¨€èªžã¯Pythonã‚’æƒ³å®šã—ã¦ã„ã¾ã™ãŒï¼Œä»–ã®è¨€èªžã«ã‚‚å¯¾å¿œã—ã¦ã„ã¾ã™
stibbar 2019/11/07
NLP

machine-learning
ãƒªãƒ³ã‚¯
KBpedia
KBpedia is an open-source knowledge graph that combines seven leading public knowledge bases into an integrated and computable structure. KBpedia has 98% coverage of Wikidata and nearly complete coverage of Wikipedia. The KBpedia distro includes its upper ontology (KKO), full knowledge graph, mappings to the major knowledge bases, and 70 largely disjoint typologies. The latest KBpedia release is v
stibbar 2018/06/13
NLP

KBPedia

wikipedia
ãƒªãƒ³ã‚¯
YAGO (database) - Wikipedia
YAGO (Yet Another Great Ontology) is an open source[3] knowledge base developed at the Max Planck Institute for Informatics in SaarbrÃ¼cken. It is automatically extracted from Wikidata and Schema.org. YAGO4, which was released in 2020, combines data that was extracted from Wikidata with relationship designators from Schema.org.[4] The previous version of YAGO, YAGO3, had knowledge of more than 10 m
stibbar 2018/06/13
NLP

YAGO

wikipedia
ãƒªãƒ³ã‚¯
ã“ã‚Œã‚‚ã‚ã‚‹æ„å‘³Deep Learningï¼ŒRecurrent Neural Network Language Modelã®è©± [MLAC2013_9æ—¥ç›®] â€” KiyuHub
ã“ã‚Œã‚‚ã‚ã‚‹æ„å‘³Deep Learningï¼ŒRecurrent Neural Network Language Modelã®è©± [MLAC2013_9æ—¥ç›®] ã“ã®æŠ•ç¨¿ã¯ Machine Learning Advent Calendar 2013 ã®9æ—¥ç›®ã®è¨˜äº‹ã§ã™ï¼Ž æœ¬æ—¥ã¯ï¼ŒNeural Network(NN)ã‚’åˆ©ç”¨ã—ãŸLanguage Model(LM)ï¼Œã¤ã¾ã‚ŠNeural Network Language Model(NNLM)ã®ä¸€ç¨®ã§ã‚ã‚Šï¼Œ Recurrent Neural Network(RNN)ã‚’ä½¿ã£ãŸRecurrent Neural Network Language Model(RNNLM)ã«ã¤ã„ã¦èª¬æ˜Žã—ã¾ã™ï¼Ž RNNLMã«è¡Œãå‰ç½®ãã¨ã—ã¦ã–ã£ãã‚Šã„ãã¤ã‹ç”¨æ„ã—ãŸã®ã§ï¼ŒçŸ¥ã£ã¦ãã†ãªäººã¯å…ˆã¸é€²ã‚“ã§ãã ã•ã„ï¼Ž å‰ç½®ãã‚’å…¥ã‚Œã¦ã„ã‚‹ã†ã¡ã«é•·ããªã‚Šã™ãŽãŸ...ï¼Ž ã—ã‹ã‚‚ï¼Œãã‚‚ãã‚‚NNã®èª¬æ˜Ž
stibbar 2018/06/13
RNN

RNNLM

NLP

Deep-learning
ãƒªãƒ³ã‚¯
WSC 2006: Lucca (Italy) â€“ World Puzzle Federation
stibbar 2018/06/13
WSC

workshop

NLP
ãƒªãƒ³ã‚¯
GitHub - asgordon/TriangleCOPA: One hundred challenge problems for logical formalizations of commonsense psychology
The term COPA is an acronym for "Choice of Plausible Alternatives," referring to a style of question for use in evaluating commonsense reasoning systems. Each question consists of a given premise and question, and two plausible alternatives - one of which is consistently selected by people. TriangleCOPA consists of 100 COPA questions in the domain of the Heider-Simmel film. Each question was valid
stibbar 2018/06/13
COPA

NLP

github
ãƒªãƒ³ã‚¯
è‡ªç„¶è¨€èªžå‡¦ç†ã®ãƒˆãƒƒãƒ—ã‚«ãƒ³ãƒ•ã‚¡ãƒ¬ãƒ³ã‚¹ - æ¦è”µé‡Žæ—¥è¨˜
NAACL HLT 2010 ã® accepted papers ã®ãƒªã‚¹ãƒˆãŒä»Šæ—¥å‡ºãŸã€‚ çŸ¥ã£ã¦ã„ã‚‹åå‰ãŒã‚ã‚‹ã‹ãªã¨æ€ã£ãŸãŒã€æ—¥æœ¬äººã²ã¨ã‚Šã‚‚ã„ãªã„ã‚ˆã†ã ã€‚ (è¨‚æ£) æ¾æœ¬ç ”å…ˆè¼©ã® tetsu-na ã•ã‚“ãŒé€šã£ã¦ã„ã‚‹ã‚ˆã†ã ã€‚@cacahoã•ã‚“æƒ…å ±ã©ã†ã‚‚ã‚ã‚ŠãŒã¨ã†ã”ã–ã„ã¾ã™! MSR æ™‚ä»£ã®ã‚¤ãƒ³ã‚¿ãƒ¼ãƒ³åŒæœŸã¨ã‹ãŸãã•ã‚“ã„ã‚‹ã€‚ã‚„ã¯ã‚Šã‚«ãƒªãƒ•ã‚©ãƒ«ãƒ‹ã‚¢é–‹å‚¬ã ã¨æ°—åˆã„å…¥ã‚Œã¦å‡ºã—ã¦ãã‚‹ã‹ã‚‰ã ã‚ã†ã‹â€¦â€¦ã€‚ä»–ã«ã‚‚å‡ºã—ã¦ã„ã‚‹ã¨èžã„ãŸäººãŒé€šã£ã¦ã„ãªã„ã®ã‚’è¦‹ã‚‹ã¨ã€ã‹ãªã‚ŠåŽ³ã—ã‹ã£ãŸæ¨¡æ§˜ã€‚ä»¥å‰ masayu-a ã•ã‚“ãŒæ—¥è¨˜ã§(ç¾åœ¨ã¯ã¦ãªãƒ€ã‚¤ã‚¢ãƒªãƒ¼ã«å¼•ã£è¶Šã—ãŸã‚ˆã†ã ãŒ) NAACL > ACL >> (è¶Šãˆã‚‰ã‚Œãªã„å£) >> CoNLL > EM NLP >> (è¶Šãˆã‚‰ã‚Œãªã„å£) >> COLING-ACL = ACL-IJCNLP > COLING(ICCL) = >> (è¶Šãˆã‚‰ã‚Œãªã„å£) >> EACL > IJCNLP
stibbar 2018/06/13
conference

EMNLP

NLP
ãƒªãƒ³ã‚¯
PDTB: Penn Discourse Treebank
The Penn Discourse Treebank Project is an NSF funded project, supported by NSF grants: IIS-14-22186 (2014-2017) IIS-14-21067 (2014-2017) CNS-10-59353 (2011-2013) IIS-07-05671 (2007-2012) CNS-02-24417 (2002-2006) The PDTB 3.0 corpus was released on March 15, 2019 through the Linguistic Data Consortium. N.B. The corpus was updated on February 4, 2020, to include the final versions of two files of to
stibbar 2018/06/13
PDTB

NLP

Treebank
ãƒªãƒ³ã‚¯
CoNLL 2018 | CoNLL
October 31 -Â November 1, 2018 Brussels, Belgium CoNLL is a top-tier conference, yearly organized byÂ SIGNLLÂ (ACL's Special Interest Group on Natural Language Learning). This year, CoNLL will be colocated with EM NLP 2018 in Brussels, Belgium. Program The program is now available! The list of accepted papers, call for papersÂ and the submission link for CoNLL 2018 areÂ still accessible. Best Paper Awar
stibbar 2018/06/13
NLP

Conference

CoNLL
ãƒªãƒ³ã‚¯
The CCG site
Combinatory Categorial Grammar (CCG) is an efficiently parseable, yet linguistically expressive grammar formalism. It has a completely transparent interface between surface syntax and underlying semantic representation, including predicate-argument structure, quantification and information structure. This site is a intended as a collection of links to CCG related projects at the University of Edin
stibbar 2018/06/13
https://qiita.com/q-ikawa/items/cf1bb593185333d88d66

NLP

CCG
ãƒªãƒ³ã‚¯
æ—¥æœ¬èªžãƒ•ãƒ¬ãƒ¼ãƒ ãƒãƒƒãƒˆ â€“ An online Japanese lexicon based on Frame Semantics
æ—¥æœ¬èªžãƒ•ãƒ¬ãƒ¼ãƒ ãƒãƒƒãƒˆã«ã¤ã„ã¦ æ—¥æœ¬èªžè©±è€…ãŒæŒã£ã¦ã„ã‚‹æ—¥æœ¬èªžèªžå½™ã«ã¤ã„ã¦ã®çŸ¥è˜ã¯ä¸€ä½“ã©ã®ã‚ˆã†ãªã‚‚ã®ã ã‚ã†ã‹ï¼Ÿç§é”ã®ãƒ—ãƒã‚¸ã‚§ã‚¯ãƒˆã§ã¯ã€ã„ã‚ã‚†ã‚‹è¾žæ›¸ã®å®šç¾©æ–‡ã®ã¿ãªã‚‰ãšä¸€èˆ¬å¸¸è˜ã‚„ã•ã‚‰ã«ç™¾ç§‘äº‹å…¸çš„çŸ¥è˜ã‚’ã‚‚å«ã‚“ã ã‚‚ã®ã€ã¨è€ƒãˆã¦ã„ã¾ã™ã€‚ç‰¹ã«ã€è©±è€…ãŒã‚ã‚‹èªžã®æ„å‘³ã‚’ç†è§£ã™ã‚‹ã«ã¯ãã®èªžã®èªçŸ¥çš„ãªèƒŒ æ™¯åŸºç›¤ï¼ˆãƒ•ãƒ¬ãƒ¼ãƒ ï¼‰ã‚’è€ƒæ…®ã™ã‚‹å¿…è¦ãŒã‚ã‚‹ã€ã¨ã„ã†ã®ãŒç§é”ã®ä»®èª¬ã§ã™ã€‚ãã—ã¦ã€ãã®ã‚ˆã†ãªæ—¥æœ¬èªžèªžå½™ã«é–¢ã™ã‚‹ãƒã‚¤ãƒ†ã‚£ãƒ–ã‚¹ãƒ”ãƒ¼ã‚«ãƒ¼ã®çŸ¥è˜ã‚’èªžå½™é …ç›®ã”ã¨ã« å®šç¾©ã—è±Šå¯Œãªä¾‹æ–‡ã‚’ã¤ã‘èªžå½™æƒ…å ±è³‡æºã¨ã—ã¦æ•´ç†ã—ã¦è¡Œã“ã†ã€ã¨ã„ã†ã®ãŒãƒ—ãƒã‚¸ã‚§ã‚¯ãƒˆã®ç©¶æ¥µã®ç›®æ¨™ã§ã™ã€‚æœ€çµ‚çš„ã«ã¯è¾žæ›¸ã¨ã‚·ã‚½ãƒ¼ãƒ©ã‚¹ã®æ©Ÿèƒ½ã‚’å…¼ãå‚™ãˆãŸã€æ—¥æœ¬èªžå¦ç¿’è€…ã‚„è‡ªç„¶è¨€èªžå‡¦ç†ã‚·ã‚¹ãƒ†ãƒ ã«å½¹ç«‹ã¤ã‚ªãƒ³ãƒ©ã‚¤ãƒ³èªžå½™æƒ…å ±è³‡æºã®æ§‹ç¯‰ã‚’ã‚ã–ã—ã¦ã„ã¾ã™ã€‚ ã“ã®èªžå½™æƒ…å ±è³‡æºã‚’ç§é”ã¯ã€Œæ—¥æœ¬èªžãƒ•ãƒ¬ãƒ¼ãƒ ãƒãƒƒãƒˆã€ï¼ˆJapanese FrameNetï¼ŒJFNï¼‰ ã¨å‘¼ã‚“ã§ã„ã¾ã™ã€‚ç¾åœ¨ã‚¢ãƒ¡ãƒªã‚«ã§æ§‹ç¯‰ã•ã‚Œã¤ã¤ã‚ã‚‹
stibbar 2018/06/13
framenet

NLP
ãƒªãƒ³ã‚¯
The Proposition Bank (PropBank)
stibbar 2018/06/13
NLP

Treebank
ãƒªãƒ³ã‚¯
Abstract Meaning Representation (AMR)
stibbar 2018/06/13
AMR

NLP
ãƒªãƒ³ã‚¯
è‡ªç„¶è¨€èªžå‡¦ç†ç”¨èªžã¨è§£æžå™¨ã¾ã¨ã‚ - Qiita
>>> import MeCab >>> tagger = MeCab.Tagger('-Owakati') >>> print tagger.parse('ã«ã‚ã«ã¯ã«ã‚ã«ã‚ã¨ã‚ŠãŒã„ã‚‹') ã«ã‚ ã« ã¯ã«ã‚ ã«ã‚ã¨ã‚Š ãŒ ã„ã‚‹ # ã¡ã‚‡ã£ã¨é–“é•ã£ã¦ã‚‹ JUMAN æœ¬ã‚·ã‚¹ãƒ†ãƒ ã¯ï¼Œè¨ˆç®—æ©Ÿã«ã‚ˆã‚‹æ—¥æœ¬èªžã®è§£æžã®ç ”ç©¶ã‚’ç›®æŒ‡ã™å¤šãã®ç ”ç©¶è€…ã«å…±é€šã«ä½¿ãˆã‚‹å½¢æ…‹ç´ è§£æžãƒ„ãƒ¼ãƒ«ã‚’æä¾›ã™ã‚‹ãŸã‚ã«é–‹ç™ºã•ã‚Œã¾ã—ãŸï¼Žãã®éš›, å¦æ ¡æ–‡æ³•ãŒè¨ˆç®—æ©Ÿå‘ãã§ã¯ãªã„ã¨ã„ã†å•é¡Œã‚’è€ƒæ…®ã—ï¼Œä½¿ç”¨è€…ã«ã‚ˆã£ã¦æ–‡æ³•ã®å®šç¾©ï¼Œå˜èªžé–“ã®æŽ¥ç¶šé–¢ä¿‚ã®å®šç¾©ãªã©ã‚’å®¹æ˜“ã«å¤‰æ›´ã§ãã‚‹ã‚ˆã†ã«é…æ…®ã—ã¾ã—ãŸ JUMAN - KUROHASHI-KAWAHARA LAB å˜ãªã‚‹å½¢æ…‹ç´ è§£æžã ã‘ã§ãªãã€Wikipediaã‹ã‚‰æŠ½å‡ºã—ãŸè¾žæ›¸ã‚„ã‚¦ã‚§ãƒ–ã‹ã‚‰ç²å¾—ã—ãŸè‡ªå‹•è¾žæ›¸ã«ã‚ˆã‚‹çµæžœãŒä»˜ä¸Žã•ã‚Œã‚‹ãŸã‚ã€MeCabã®è¾žæ›¸ã§ã¯æœªçŸ¥èªžã«ãªã£ã¦ã—ã¾ã†ã‚ˆã†ãªå›ºæœ‰åè©žãªã©ã«å¼·ã„ã€‚ã¾ãŸã€å‹•è©žã«å¯¾ã—
stibbar 2018/06/13
qiita

NLP

corpus
ãƒªãƒ³ã‚¯
ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ - Wikipedia
ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ï¼ˆè‹±: Treebankï¼‰ã¯ã€ã‚³ãƒ¼ãƒ‘ã‚¹ã®ä¸€ç¨®ã§ã‚ã‚Šã€å„æ–‡ã«çµ±èªžæ§‹é€ ã®æ³¨é‡ˆãŒä»˜ä¸Žã•ã‚Œã¦ã„ã‚‹ã‚‚ã®ã§ã‚ã‚‹ã€‚çµ±èªžæ§‹é€ ã¯ä¸€èˆ¬ã«æœ¨æ§‹é€ ã§è¡¨ã•ã‚Œã‚‹ã“ã¨ãŒå¤šã„ãŸã‚ã€ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ã¨å‘¼ã°ã‚Œã‚‹ã€‚ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ã¯ã‚³ãƒ¼ãƒ‘ã‚¹è¨€èªžå¦ã§æ–‡æ³•çš„ç¾è±¡ã®ç ”ç©¶ã«ä½¿ã‚ã‚Œã‚‹ä»–ã€è¨ˆç®—è¨€èªžå¦ã§ã®æ§‹æ–‡è§£æžå™¨ã®è©•ä¾¡ã‚„è¨“ç·´ã«ä½¿ã‚ã‚Œã‚‹ã€‚ ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ã¯æ—¢å˜ã®ä½•ã‚‰ã‹ã®ã‚¿ã‚°ãŒæ—¢ã«ä»˜ã„ãŸã‚³ãƒ¼ãƒ‘ã‚¹ã‚’ä½¿ã£ã¦ä½œæˆã•ã‚Œã‚‹ã“ã¨ãŒå¤šã„ã€‚å¾“ã£ã¦ã€ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ã¯æ„å‘³è«–çš„ã«ä½•ã‚‰ã‹ã®æ‹¡å¼µã‚’ã—ãŸã‚Šã€ä»–ã®è¨€èªžå¦çš„æƒ…å ±ã‚’ä»˜ä¸Žã—ãŸã‚Šã—ã¦ã„ã‚‹ã€‚ ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ã¯ã€è¨€èªžå¦è€…ãŒæ–‡ä¸€ã¤ä¸€ã¤ã«æ³¨é‡ˆã‚’åŠ ãˆã¦ã„ã£ã¦æ§‹ç¯‰ã™ã‚‹ã“ã¨ã‚‚ã‚ã‚‹ã—ã€åŠè‡ªå‹•çš„ã«æ§‹æ–‡è§£æžå™¨ã§è§£é‡ˆã—ã€ãã®çµæžœã‚’è¨€èªžå¦è€…ãŒãƒã‚§ãƒƒã‚¯ã™ã‚‹ã¨ã„ã†æ–¹å¼ã‚‚ã‚ã‚‹ã€‚ ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ã¯ç‰¹å®šã®è¨€èªžå¦çš„ç†è«–ï¼ˆä¾‹ãˆã°HPSGï¼‰ã«åŸºã¥ã„ãŸæ–‡æ³•çš„æ³¨é‡ˆã‚’ä»˜ä¸Žã—ã¦ã„ã‚‹å ´åˆã‚‚ã‚ã‚‹ãŒã€å¤šãã®ãƒ„ãƒªãƒ¼ãƒãƒ³ã‚¯ã¯ç‰¹å®šã®ç†è«–ã«ä¾å˜ã—ãªã„ã‚ˆã†ã«ã—ã¦ã„ã‚‹ã€‚ã—ã‹ã—ãªãŒã‚‰ã€
stibbar 2018/06/13
https://www.ling.upenn.edu/courses/Fall_2003/ling001/penn_treebank_pos.html

NLP

Wikipedia

Treebank
ãƒªãƒ³ã‚¯
1 2 æ¬¡ã®ãƒšãƒ¼ã‚¸