Google Neural Machine Translation

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search

Google Neural Machine Translation (GNMT) is a neural machine translation (NMT) system developed by Google and introduced in November 2016, that uses an artificial neural network to increase fluency and accuracy in Google Translate.[1][2][3]

GNMT improves on the quality of translation by applying an example based (EBMT) machine translation method in which the system "learns from millions of examples".[2] GNMT's proposed architecture of system learning was first tested on over a hundred languages supported by Google Translate.[2] With the large end-to-end framework, the system learns over time to create better, more natural translations.[1] GNMT is capable of translating whole sentences at a time, rather than just piece by piece.[1] The GNMT network can undertake interlingual machine translation by encoding the semantics of the sentence, rather than by memorizing phrase-to-phrase translations.[2][4]

History[edit]

The Google Brain project was established in 2011 in the "secretive Google X research lab"[5] by Google Fellow Jeff Dean, Google Researcher Greg Corrado, and Stanford University Computer Science professor Andrew Ng.[6][7][8] Ng’s work has led to some of the biggest breakthroughs at Google and Stanford.[5]

In September 2016, a research team at Google announced the development of the Google Neural Machine Translation system (GNMT) and by November Google Translate began using neural machine translation (NMT) in preference to its previous statistical methods (SMT)[1][9][10][11] which had been used since October 2007, with its proprietary, in-house SMT technology.[12][13]

Google Translate's NMT system uses a large artificial neural network capable of deep learning.[1][2][3] By using millions of examples, GNMT improves the quality of translation,[2] using broader context to deduce the most relevant translation. The result is then rearranged and adapted to approach grammatically based human language.[1] GNMT's proposed architecture of system learning was first tested on over a hundred languages supported by Google Translate.[2] GNMT did not create its own universal interlingua but rather aimed at commonality found in between many languages, considered to be of more interest to psychologists and linguists than to computer scientists.[14] The new translation engine was first enabled for eight languages: to and from English and French, German, Spanish, Portuguese, Chinese, Japanese, Korean and Turkish in 2016.[15] In March 2017, three additional languages were enabled: Russian, Hindi and Vietnamese along with Thai for which support was added later.[16][17] Support for Hebrew and Arabic was also added with help from the Google Translate Community in the same month.[18] In mid April 2017 Google Netherlands announced support for Dutch and other European languages related to English.[19] Further support was added for nine Indian languages, viz. Hindi, Bengali, Marathi, Gujarati, Punjabi, Tamil, Telugu, Malayalam and Kannada at the end of April 2017.[20]

Languages Supported by GNMT[edit]

This is a list of language translation pairs supported by Google Translate's Neural Machine Translation (NMT) model. As of July 2017 all languages currently only support translation to and from English:[21]

Language Pair Language Codes
Afrikaans <-> English af <-> en
Albanian <-> English sq <-> en
Amharic <-> English am <-> en
Arabic <-> English ar <-> en
Armenian <-> English hy <-> en
Azerbaijani <-> English az <-> en
Basque <-> English eu <-> en
Bengali <-> English bn <-> en
Bosnian <-> English bs <-> en
Bulgarian <-> English bg <-> en
Catalan <-> English ca <-> en
Cebuano <-> English ceb <-> en
Chinese (Simplified) <-> English zh-CN * <-> en
Chinese (Traditional) <-> English zh-TW <-> en
Corsican <-> English co <-> en
Croatian <-> English hr <-> en
Czech <-> English cs <-> en
Danish <-> English da <-> en
Dutch <-> English nl <-> en
Esperanto <-> English eo <-> en
Estonian <-> English et <-> en
Finnish <-> English fi <-> en
French <-> English fr <-> en
Frisian <-> English fy <-> en
Galician <-> English gl <-> en
Georgian <-> English ka <-> en
German <-> English de <-> en
Greek <-> English el <-> en
Gujarati <-> English gu <-> en
Haitian Creole <-> English ht <-> en
Hausa <-> English ha <-> en
Hawaiian <-> English haw <-> en
Hebrew <-> English iw <-> en
Hindi <-> English hi <-> en
Hmong <-> English hmn <-> en
Hungarian <-> English hu <-> en
Icelandic <-> English is <-> en
Igbo <-> English ig <-> en
Indonesian <-> English id <-> en
Irish <-> English ga <-> en
Italian <-> English it <-> en
Japanese <-> English ja <-> en
Javanese <-> English jw <-> en
Kannada <-> English kn <-> en
Kazakh <-> English kk <-> en
Khmer <-> English km <-> en
Korean <-> English ko <-> en
Kurdish <-> English ku <-> en
Lao <-> English lo <-> en
Latvian <-> English lv <-> en
Lithuanian <-> English lt <-> en
Luxembourgish <-> English lb <-> en
Macedonian <-> English mk <-> en
Malagasy <-> English mg <-> en
Malay <-> English ms <-> en
Malayalam <-> English ml <-> en
Maltese** <- English mt <- en
Maori <-> English mi <-> en
Marathi <-> English mr <-> en
Mongolian <-> English mn <-> en
Nepali <-> English ne <-> en
Norwegian <-> English no <-> en
Nyanja (Chichewa) <-> English ny <-> en
Pashto <-> English ps <-> en
Persian <-> English fa <-> en
Polish <-> English pl <-> en
Portuguese (Portugal, Brazil) <-> English pt <-> en
Punjabi <-> English pa <-> en
Romanian <-> English ro <-> en
Russian <-> English ru <-> en
Samoan <-> English sm <-> en
Scots Gaelic <-> English gd <-> en
Serbian <-> English sr <-> en
Sesotho <-> English st <-> en
Shona <-> English sn <-> en
Sindhi <-> English sd <-> en
Sinhala (Sinhalese) <-> English si <-> en
Slovak <-> English sk <-> en
Slovenian <-> English sl <-> en
Somali <-> English so <-> en
Spanish <-> English es <-> en
Swahili <-> English sw <-> en
Swedish <-> English sv <-> en
Tagalog (Filipino) <-> English tl <-> en
Tajik <-> English tg <-> en
Tamil <-> English ta <-> en
Telugu <-> English te <-> en
Thai <-> English th <-> en
Turkish <-> English tr <-> en
Ukrainian <-> English uk <-> en
Urdu <-> English ur <-> en
Uzbek <-> English uz <-> en
Vietnamese <-> English vi <-> en
Welsh <-> English cy <-> en
Xhosa <-> English xh <-> en
Yiddish <-> English yi <-> en
Yoruba <-> English yo <-> en
Zulu <-> English zu <-> en

Zero-shot translation[edit]

The GNMT system is said to represent an improvement over the former Google Translate in that it can handle "zero-shot translation", that is it directly translates one language into another (for example, Japanese to Korean).[2] Google Translate previously first translated the source language into English and then translated the English into the target language rather than translating directly from one language to another.[4]

See also[edit]

References[edit]

  1. ^ a b c d e f Barak Turovsky (November 15, 2016), "Found in translation: More accurate, fluent sentences in Google Translate", Google Blog, retrieved January 11, 2017 
  2. ^ a b c d e f g h Mike Schuster, Melvin Johnson, and Nikhil Thorat (November 22, 2016), "Zero-Shot Translation with Google's Multilingual Neural Machine Translation System", Google Research Blog, retrieved January 11, 2017 
  3. ^ a b Gil Fewster (January 5, 2017), "The mind-blowing AI announcement from Google that you probably missed", freeCodeCamp, retrieved January 11, 2017 
  4. ^ a b Boitet, Christian; Blanchon, Hervé; Seligman, Mark; Bellynck, Valérie (2010). "MT on and for the Web" (PDF). Retrieved December 1, 2016. 
  5. ^ a b Robert D. Hof (August 14, 2014). "A Chinese Internet Giant Starts to Dream: Baidu is a fixture of online life in China, but it wants to become a global power. Can one of the world's leading artificial intelligence researchers help it challenge Silicon Valley's biggest companies?". Technology Review. Retrieved January 11, 2017. 
  6. ^ Jeff Dean and Andrew Ng (June 26, 2012). "Using large-scale brain simulations for machine learning and A.I." Official Google Blog. Retrieved January 26, 2015. 
  7. ^ "Google's Large Scale Deep Neural Networks Project". Retrieved October 25, 2015. 
  8. ^ Markoff, John (June 25, 2012). "How Many Computers to Identify a Cat? 16,000". New York Times. Retrieved February 11, 2014. 
  9. ^ Katyanna Quach (November 17, 2016), Google's neural network learns to translate languages it hasn't been trained on: First time machine translation has used true transfer learning, retrieved January 11, 2017 
  10. ^ Lewis-Kraus, Gideon (December 14, 2016). "The Great A.I. Awakening". The New York Times. Retrieved January 11, 2017. 
  11. ^ Le, Quoc; Schuster, Mike (September 27, 2016). "A Neural Network for Machine Translation, at Production Scale". Google Research Blog. Google. Retrieved December 1, 2016. 
  12. ^ Google Switches to its Own Translation System, October 22, 2007
  13. ^ Barry Schwartz (October 23, 2007). "Google Translate Drops SYSTRAN for Home-Brewed Translation". Search Engine Land. 
  14. ^ Chris McDonald (January 7, 2017), Commenting on Gil Fewster's January 5th article in the Atlantic, retrieved January 11, 2017 
  15. ^ Turovsky, Barak (November 15, 2016). "Found in translation: More accurate, fluent sentences in Google Translate". The Keyword Google Blog. Google. Retrieved December 1, 2016. 
  16. ^ Perez, Sarah (March 6, 2017). "Google's smarter, A.I.-powered translation system expands to more languages". TechCrunch. Oath Inc. 
  17. ^ Turovsky, Barak. "Higher quality neural translations for a bunch more languages". The Keyword Google Blog. Google. Retrieved March 6, 2017. 
  18. ^ Novet, Jordan (March 30, 2017). "Google now provides AI-powered translations for Arabic and Hebrew". VentureBeat. 
  19. ^ Finge, Rachid (April 19, 2017). "Grote verbetering voor het Nederlands in Google Translate" [Big improvement for Dutch in Google Translate]. Google Netherlands Blog (in Dutch). 
  20. ^ Turovsky, Barak (April 25, 2017). "Making the internet more inclusive in India". The Keyword. 
  21. ^ "Translation API Language Support". Google Cloud Platform. May 4, 2017. 

External links[edit]