Experimenting to solve cramming
When we make a decision as significant as changing the number of characters available for Tweeting, we need to know it’s right. How do we figure that out? Through a rigorous evaluation process to ensure any change — especially one at the core of who we’ve been for more than a decade – improves everyone’s experience on Twitter. We started looking at making a change to our 140-character limit by per
若者言葉「ã•ã‚ã‚‹ã€å½¢ã®ä½¿ç”¨å®Ÿæ…‹ã¨ã€ ãã®ç‰¹å¾´ã«ã¤ã„㦠- ã‚ã¨ãƒ¼ã™ãƒã‚°
（ã“ã®ã‚¨ãƒ³ãƒˆãƒªã¯ã€LiteraTech風見é¶ã«2014/9/30ã«æŽ²è¼‰ã—ãŸã‚¨ãƒ³ãƒˆãƒªã‚’å†ç·¨é›†ã—ãŸã‚‚ã®ã§ã™ã€‚） ï¼ï¼Žã¯ã˜ã‚ã«â€¦â€¦ã®å‰ã« 「１．ã¯ã˜ã‚ã«ã€ä»¥é™ã¯ã€è¬›ç¾©ã®æœŸæœ«ãƒ¬ãƒãƒ¼ãƒˆã«è‹¥å¹²ã®ä¿®æ£ã‚’åŠ ãˆãŸã‚‚ã®ã¨ãªã£ã¦ã„ã¾ã™ã€‚ã—ã‹ã—ã€æ¦‚ãæ出ã—ãŸã‚‚ã®ã¨å¤‰ã‚らãªã„ã§ã™ã€‚表やグラフãŒã‚„や見ã¥ã‚‰ããªã£ã¦ã„ã¾ã™ãŒã€ã”容赦ãã ã•ã„。ã¾ãŸã€ã‚¹ãƒžãƒ›ã ã¨è¡¨ç¤ºãŒå´©ã‚Œã‚‹æã‚ŒãŒã‚ã‚‹ã®ã§ã€PCã§ã®é–²è¦§ã‚’推奨ã—ã¾ã™ã€‚ １．ã¯ã˜ã‚ã« Twitterを使用ã—ã¦ã„ã‚‹ã¨ã€å¤§å¦ã®åŒç´šç”ŸãŒã€Œã•ã‚ã‚‹ã€ã¨ã„ã†è¨€è‘‰ã‚’使ã£ã¦ã„ã‚‹ã“ã¨ã‚’ç›®ã«ã™ã‚‹ã€‚自分ã®å‘¨ã‚Šã ã‘ã§ä½¿ã‚ã‚Œã¦ã„る表ç¾ã‹ã¨æ€ã„ã€Twitter検索ã§èª¿ã¹ã¦ã¿ã‚‹ã¨ã€åŒã˜ã‚ˆã†ã«ã€Œã•ã‚ã‚‹ã€ã‚’使ã£ã¦ã„る人ãŒè¤‡æ•°äººã„ã‚‹ã“ã¨ãŒã‚ã‹ã£ãŸã€‚ ãã“ã§ã€æœ¬ç¨¿ã§ã¯ã“れを「ã•ã‚ã‚‹ã€å½¢ã¨å付ã‘ã€Twitter上ã§ã®ä½¿ç”¨ä¾‹ã‚’見ãªãŒã‚‰ã€ã©ã†ã„ã£ãŸæ™‚ã«ã“ã®ã€Œã•ã‚ã‚‹ã€å½¢ãŒä½¿ã‚れるã®ã‹ã€ã€Œã•ã‚ã‚‹ã€å½¢ã«ã¯ã©ã®ã‚ˆã†ãªç‰¹å¾´ãŒã‚ã‚‹ã®ã‹
Evaluating language identification performance
Understanding the content of Tweets is important for many reasons: grasping a user’s interests (which in turn lets us show more relevant content), improving search, and fighting spam. There are many steps involved in a typical natural language processing pipeline, but one of the first and most fundamental steps is language identification — determining the language in which a piece of text is writt
KDD2014勉強会: Large-Scale High-Precision Topic Modeling on Twitter
2. Large-Scale High-Precision Topic Modeling on Twitterby ShuangYang, Alek Kolcz, Andy Schlaikjer, Pankaj Gupta (Twitter, Inc.) •大é‡ã«è“„ç©ã•ã‚Œã‚‹Tweetã«å¯¾ã—ã¦ã€ã‚らã‹ã˜ã‚用æ„ã—ãŸ300+種類ã®â€œãƒˆ ピック†付与å•é¡Œã‚’マルãƒãƒ©ãƒ™ãƒ«å•é¡Œã¨ã—ã¦è§£ã –ã„ã‚ゆる“Topic Model†ãªè©±ã§ã¯ãªã„ïŒ â€“æœ¬ç™ºè¡¨ã§ã¯ãƒˆãƒ”ックã®ä»£ã‚ã‚Šã«ã‚¿ã‚°ã¨è¡¨ç¾ •様々ãªå·¥å¤« –訓練データã®è‡ªå‹•ç²å¾— –高速ãªç‰¹å¾´æŠ½å‡º –閾値ã®ãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚° â€“ã‚¿ã‚°éšŽå±¤æ§‹é€ ã®åˆ©ç”¨ –ãªã© •Precision 93% (37% coverage) ã®ã‚¿ã‚°å‰²ã‚Šå½“ã¦ç²¾åº¦ã‚’é”æˆ Twitterã§å®Ÿé‹ç”¨å¯èƒ½ãªãƒ¬ãƒ™ãƒ«ã®é«˜Precisionã®ã‚¿ã‚°å‰²ã‚Šå½“ã¦æ©Ÿèƒ½ã‚’å®Ÿç¾ 1æžšæ¦‚è¦ Industry paper 2 ※R
Flight by Twitter
A lightweight, component-based JavaScript framework for assigning behavior to DOM nodes.Flight A lightweight, component-based JavaScript framework from Twitter View on GitHub Overview Flight is a lightweight, component-based JavaScript framework that maps behavior to DOM nodes. Twitter uses it for their web applications. By way of example, we've including a simple email client demo (browse the sou
NAIST 㧠twitter 言語判定ã«ã¤ã„ã¦ç™ºè¡¨ã—ã¦ãã¾ã—㟠- 木曜ä¸è¶³
5/14 ã« NAIST(奈良先端科å¦æŠ€è¡“大å¦é™¢å¤§å¦) ã«ã¦ã€Œâˆž-gramを使ã£ãŸçŸæ–‡è¨€èªžåˆ¤å®šã€ã«ã¤ã„ã¦ç™ºè¡¨ã—ã¦ãã¾ã—ãŸã€‚ç´ æ™´ã‚‰ã—ã„機会をã„ãŸã ã„ã¦ã‚ã‚ŠãŒã¨ã†ã”ã–ã„ã¾ã—ãŸã€‚ ãã®æ™‚ã®è³‡æ–™ã‚’公開ã—ã¾ã™ã€‚ 内容㯠TokyoNLP #8 や言語処ç†å¦ä¼š2012ï¼ åºƒå³¶å¸‚å¤§ã§ç™ºè¡¨ã—ãŸã‚‚ã®ã‚’ã¡ã‚‡ã£ã¨ä¸å¯§ã«ï¼‹Î±ã—ãŸæ„Ÿã˜ã€‚ Short Text Language Detection with Infinity-Gram View more presentations from Shuyo Nakatani NAIST å°ç”ºã•ã‚“ã‹ã‚‰ M1 å‘ã‘ã®ã‚»ãƒŸãƒŠãƒ¼ã§ç™ºè¡¨ã—ã¦ã¿ã¾ã›ã‚“ã‹ã¨ãŠè©±ã‚’ã„ãŸã ã„ã¦ã€ã€Œãœã²ã‚„ã‚Šã¾ã™ã€ã¨äºŒã¤è¿”事ã§ãŠå¼•ãå—ã‘ã—ãŸã‚‚ã®ã®ã€æœ¬å½“ã«ä¸è°·ã§ã„ã„ã‚“ã ã‚ã†ã‹ãƒ¼ã¨ã‚ã¨ã‹ã‚‰å¿ƒé…ã«ãªã£ãŸã®ã¯ãªã„ã—ょｗ å°ç”ºã•ã‚“ã‹ã‚‰ã¯ã€æ©Ÿæ¢°å¦ç¿’や自然言語処ç†ã‚’実応用ã§ä½¿ã£ã¦ã„るエンジニアã®ç«‹å ´ã€ã¨ã„ã†ã¨ã“ã‚ã‚’å¦ç”Ÿã•ã‚“ã«
1
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
TechCrunch | Startup and Technology News
{{#tags}}- {{label}}
{{/tags}}