September 26, 2008

More Languages in Google Translate

Google Translate added 11 new languages: Catalan, Filipino, Hebrew, Indonesian, Latvian, Lithuanian, Serbian, Slovak, Slovenian, Ukrainian and Vietnamese. Google's machine translation service now supports 35 languages and you can use it to translate text between any combination of languages.

In most cases, Google uses English as an intermediary language, so when you translate a text from Indonesian to Vietnamese, Google translates the text to English and then it translates the result to Vietnamese. You'll get the best results when one of the languages is English, since Google needs a single translation.


"Most state-of-the-art, commercial machine-translation systems in use today have been developed using a rule-based approach, and require a lot of work to define vocabularies and grammars. Our system takes a different approach: we feed the computer billions of words of text, both monolingual text in the target language, and aligned text consisting of examples of human translations between the languages. We then apply statistical learning techniques to build a translation model," explains Google.

One of the advantages of this approach is scalability: if Google finds enough parallel text to create a good translation model for a language, it will be added to Google Translate. When Google licensed Systran's technology, Google Translate was only able to translate between English and French, German, Spanish, Italian and Portuguese, but this has changed when Google developed its own translation technology. Microsoft followed suit and Windows Live Translator switched from Systran to Microsoft's machine translation system.

As Microsoft notes, it's important to keep in mind that "automatic translation enables you to understand the gist of foreign language text, but is no substitute for a professional human translator if fluency is required," at least not yet.

110 comments:

  1. I think Turkish sould be added.

    ReplyDelete
  2. Great! And I would like to see more Southeast Asian languages to be added.

    ReplyDelete
  3. Machine translation, including Google's, is not much better than it was 40+ years ago when I was studying it in college. Brute force will never solve the problem, no matter how many CPU cycles or petabytes of data you throw at it. It needs to be approached differently. Perhaps trying to mimic mental processes would work bettter, now that we can zero in on some mental activity via an MRI.

    ReplyDelete
  4. turkish definetly shoul be added `cuz world doesnt know what it misses
    lol

    ReplyDelete
  5. Google uses Spanish as an intermediary language with Catalan.
    The translation spanish-catalan-spanish is almost 99,9% perfect.
    Wow

    ReplyDelete
  6. Turkish language structure is very complicated, you always append millions different kind of addition at the end of a word.It s no like other languages, i think it is reason why they are not able to that, even though there is a lot of turkish google users than other small country users.

    ReplyDelete
  7. alhamdulillah... Indonesian language are inserted among other language.... Google rocks!!

    ReplyDelete
  8. You got Finnish, Latvian, Russian and Swedish.
    But Estonia is just in middle of them... Please add us too :)

    ReplyDelete
  9. I was delighted to discover Hebrew among the available languages today! Even though as you mention, it only "enables you to understand the gist"

    ReplyDelete
  10. Hi everyone, you are so great people to offer us language tools for free, thanx!! But could you please add the Frisian [=Fries = Frysk] language, because you have got Dutch, so Frisian shuold be present too.
    And what about Saqartvelo, Mongolian, Cambodjan, Amharic, Aramaic, Syraic??
    I would be very pleased to see these languages on GOOGLE TRANSLATE... : )

    ReplyDelete
  11. Thank's for Catalan inclusion ;-)

    ReplyDelete
  12. http://www.blogger.com/captcha?type=IMAGE&captchaKey=3xrk1r9xf93p

    ReplyDelete
  13. I must say that the translation from Hebrew to English in the example above shows just how much more work is needed in order to perfect computerized translations.

    Take for instance the airline KLM that appears in the article. Since the Hebrew letters do not match the latin ones, the letters are sounded out (Kay - El - Em) and the computerized translation then translates the meaning of the words El and Em (To and If).

    There are also 3 instances in-which words are that weren't recognized were transposed phonetically (poorly too)... The words for Intended, Holland and Citizen.

    ReplyDelete
  14. "Google Translate in Latvian"

    http://libeks.blogspot.com/2008/09/google-translate-in-latvian.html

    ReplyDelete
  15. Any idea when these new languages will be made into bots for google talk?

    ReplyDelete
  16. My mother tongue Tamil is still missing. When it will be available :(

    ReplyDelete
  17. never heard catalan and serbian before lol.

    ReplyDelete
  18. ''aqartvelo, Mongolian, Cambodjan, Amharic, Aramaic, Syraic'' oh my god? human languages?

    ReplyDelete
  19. Nice step taken by Google.But wat abt Bengali language? is tre any specific translator which can translate Bengali language into English?

    ReplyDelete
  20. I think are great . but where is urdu ?

    ReplyDelete
  21. Did anybody compare the Google's translation with other translation tools (especilly, free tools)? better or worse?

    ReplyDelete
  22. I don't know what Google is doing but when did word Slovenian become translated to Croatia?

    http://209.85.135.104/translate_c?hl=en&sl=sl&tl=en&u=http://www.slo-foto.net/reviews-137.html&usg=ALkJrhgFcpjGBsmbW0nGRnm0VT-4o_N4IQ

    Original text:
    Spletne strani slovenskih fotografskih društev

    Translated (wrong):
    Websites of the Croatian photographic societies

    Translated (correct):
    Websites of the Slovenian photographic societies

    ReplyDelete
  23. @Anonymous:
    You must understand that Google uses statistical machine translation, so the results depend a lot on the parallel texts used to create language models.

    The word "slovenskih" is translated in different ways, depending on the context. Try to translate "slovenskih" at translate.google.com and you'll get the right translation. Unfortunately, Croatian and Slovenian are used in similar contexts and Google couldn't find the right correspondence. As Google obtains more parallel texts, the translation's quality will improve.

    ReplyDelete
  24. Thanks for the explanation Alex!

    It's not nice to see that the name of your country is translated to the name of your neighbor ;)

    But otherwise I am very very happy that Google has finally implanted our language :)

    ReplyDelete
  25. Turkish please..it is more important a lot of language in there....250-300 millions of people speak Turkish.....it is very important fou us and you...forward

    ReplyDelete
  26. Google must do Turkish Translate.

    ReplyDelete
  27. Yay for latvian translation!

    ReplyDelete
  28. Google, terimakasih atas masuknya Bahasa Indonesia dalam service yang disediakan. Semoga di masa yang akan datang bahasa-bahasa lain dari negara saya bisa pula dimasukkan, seperti Bahasa Jawa, Bahasa Batak, Bahasa Sunda dan lain-lain. Semoga Google semakin maju!

    For those who do not speak English, you can translate my comment above by translating it from Indonesian to English.

    ReplyDelete
  29. thanks for indonesian

    ReplyDelete
  30. Many thx for Ukrainian! it's going to make a fantastic difference to me. Cheers! mm

    ReplyDelete
  31. How can I help to make Faroese a part of Google Translate?

    ReplyDelete
  32. I agree. Definitely we need turkish translation.

    ReplyDelete
  33. What about Estonian? I would love to see Estonian added too.

    ReplyDelete
  34. just to let you know, I made a little widget with flags for translation. the code is creative commons and can be used by anyone, no link required (though one is requested).

    the page is in a language you may or may not know about :)

    ReplyDelete
  35. where is turkish translate!?? even latvian translate exists but turkish not! thats not fair..

    ReplyDelete
  36. Google translates Filipino language erroneously. The entire meaning changed drammatically when I tried to translate a tagalog paragraph into English.
    I tried changing my gmail language settings into filipino, and the same thing happened. It is actually funny and irritating at the same time that the words and terms that are similar in meaning are interchanged and sometimes terms are used in the wrong context.

    ReplyDelete
  37. Why not add African languages as well? Such as... Swahili, Zulu or Siswati?

    ReplyDelete
  38. i want to translate turkish languange can you add turkish in google translation because i am always using google translate

    ReplyDelete
  39. What the hell is wrong with you guys? Google Translate isn't obligated to include EVERY SINGLE language out there, just because it happens to be YOUR native tongue.

    I can sort of understand Turkish, since it's the 23rd-most-spoken language (which really isn't that high, but if Google already has 35 languages so far, then it makes logistical sense), but those of you suggesting (or even DEMANDING) all these obscure languages that have very little global impact are just being crazy, if not selfish.

    Google has other priorities, you know.

    ReplyDelete
  40. Mr. Matt,
    please look at Czech which is 65th most spoken language in the world. It isn't about the popularity but something else it seems.

    Regards,

    ReplyDelete
  41. Our 80 Milyon population there, we want turkish.
    Small country there
    turkish translate............
    We are deserveing this

    ReplyDelete
  42. I hope Malay language can be add...

    ReplyDelete
  43. Where is Turkish? One of the world's most spoken languages.

    ReplyDelete
  44. Thai should be included as Thailand is now a major tourist centre in Asia, more so in fact than China. It is also a manufacturing nation.

    ReplyDelete
  45. Noticed the default language beginning in the drop down is Spanish. How can I change this to Chinese or another language?

    ReplyDelete
  46. WE WANT TURKISH!!!! İMMEDİATELY

    ReplyDelete
  47. We want turkish....You aren't doing on purpose :(

    ReplyDelete
  48. Hey "GOOGLE" do you know how many Albanian users wants to use your translating services???!
    Pls add ALBANIAN as fast you can!!!!!!!!!!!??????????????????

    ReplyDelete
  49. I would LOVE to see Estonian in there (:

    ReplyDelete
  50. Urdu should be added!!!, v dont understand english or anyother language :)

    vote for urdu :D

    ReplyDelete
  51. There is no logic some of a languages you have two times, such as the Serbian and Croatian-(it is the same language),oter side , some languages simply do not exsist.Example, Macedonian, Turkish, Albanian...
    I suggest to remove the Greek language and insert the Macedonian , because the entire so-called Greece is Makedonija.Thanks!

    Krste Krstevski,Garfield,NJ USA

    ReplyDelete
  52. KRSTE KRSTEVSKI, Garfield, NJ, USA

    We hope that soon will be introduced and Macedonian language.

    Thank you!

    ReplyDelete
  53. turkish lang is needed...

    ReplyDelete
  54. Krste Krstevski
    Bitola-Dimce Lahcanski-47
    Republik of Macedonia

    it is necessary to enter the Macedonian language, because half of Macedonia is in the United States, Australia, Canada and many other countries around the world.Magotity of them use this services.If this service is because of practical resons, please enter the Macedonian language.

    THANK YOY!

    ReplyDelete
  55. Jas mislam deka treba da go vnesite i MACEDONIAN LANGUAGE,zatoa sto
    imame potreba od toa.Zatoa sto nie Makedoncite (nekoi od nas)se vo ista sostojba kako Vie sega sto ne go razbirate ova.
    Please M A C E D O N I A N language.

    ReplyDelete
  56. Please enter and Macedonian language Thanks!

    Simeon Sopov, Bitola, Macedonia.

    ReplyDelete
  57. Ihope soon- M A C E D O N I A N language will be introducet.


    Anka from Bitola-Macedonia.

    ReplyDelete
  58. Emilia

    Please for Macedonial langauge.Thanks.

    ReplyDelete
  59. Daniel-NJ, USA

    please I N T R O D U C E M A C E D O N I A N L A N G U A G E !

    Absolutely unusual is to be exercised Translation from English to English, Spanihs to Spanich ...... But in Googol it is possible for egzample: Serbian to Croatian which is a same language.
    Please to introduce and MAKEDONIAN.
    Thanks.

    ReplyDelete
  60. Zorica:

    It is necessary to enter the MACEDONIAN LANGUAGE. Thanks!

    ReplyDelete
  61. Margorie:
    Macedonian langguage .Thanks.

    ReplyDelete
  62. I have a great need to be
    Macedonian language here together with other languages.
    Thanks!

    ReplyDelete
  63. M A C E D O N I A N language shoot be here together with other languages.
    Mu blagodarime na Googol ako go stori toa.

    ReplyDelete
  64. Where's the Turkish :O

    ReplyDelete
  65. Hi everyone,

    I think we do need before any of the languages in Google list the Assyrian (Syriac/Aramaic) language because it's the most important language in the whole World.

    ReplyDelete
  66. I think the irish language should be added

    ReplyDelete
  67. Yeah, Absolutely I agree with my Turkish folks!!!
    You can speak Turkish from Anatolia to Mongolia!!
    Well, I accept about the different grammatics))

    ReplyDelete
  68. People keep saying they want Turkish language added, yeah I want that too. However you miss some little detail, which in long terms become a BIG problem. Turkish does not have the same grammar structure as most of the languages included in Google Translate. And if you still want to believe Google is ignoring you, just go and type a Turkish sentence into Google Translate and set to auto-detect language. It will tell you "Turkish to English is not available yet". Which means the database is being prepared (or some idiots can think "YaY, Computer can speak TURKISH!!!). Either way sooner or later we will get Turkish. But give them some time or it will be very deficient and unusable as in fluency and grammar wise.

    ReplyDelete
  69. Macedonian langauge -please!!!!

    ReplyDelete
  70. Für uns ist viel MAZEDONISCH Sprache.

    Danke!

    ReplyDelete
  71. Bitte für die mazedonische Sprache!

    Danke!

    ReplyDelete
  72. i think there should be Turkish added!!!

    ReplyDelete
  73. Si prega di lingua MACEDONE ...

    ReplyDelete
  74. Μακεδονική γλώσσα είναι είναι απαραίτητο να εισαχθεί.

    ReplyDelete
  75. turkish sould b added

    ReplyDelete
  76. How long will we wait for Google to introduce Macedonia language?!

    ReplyDelete
  77. (MACEDONIAN Language)

    ReplyDelete
  78. (Μακεδονική),(MACEDONIAN Language).

    ReplyDelete
  79. MAZEDONISCHE SPRACHE!!!

    ReplyDelete
  80. Comment is not empty :
    Hope that soon will be introduced and Macedonian language.

    ReplyDelete
  81. I hope soon going to see and Macedonian language.

    Thank you!

    ReplyDelete
  82. From MACEDONIANS!
    Happy New Year to Google.

    ReplyDelete
  83. Thanks for Macedonian langguage .

    ReplyDelete
  84. i thin that alabanian language is missing

    ReplyDelete
  85. Mhm, no Estonian.

    I guess Estonian is a bit too complicated for you to handle.

    ReplyDelete
  86. svp il faudrait ajouter le turc

    ReplyDelete
  87. Yes, What about Estonian?.
    Is this langue to hard for you guys?? Or we are so unimportant??

    ReplyDelete
  88. turkish please.....

    ReplyDelete
  89. albanian please please we need so muchh :$:$ please

    ReplyDelete
  90. Macedonian is not a language dude.... If you are macedonian, then greek should be fine for you!!!
    Else, try albanian, servian or bulgarian something!!

    ReplyDelete
  91. while there are chinese japanese and even arabic,it musnt be so complicated to add the Turkish language at the translator i guess..

    ReplyDelete
  92. I needed Estonian but it isn't there. The translation is often using the wrong word so I wouldn't use it to translate text for publication, it's only good for reading yourself to get some idea of what is said.

    ReplyDelete
  93. The flag for Macedonia,the post for thr grecs.

    ReplyDelete
  94. thanks for turkish
    but i cant think of that translations wont appropriate

    turkish is very difficult to learn and/or translate any other languages
    an example
    bahtsız bedeviyi çölde kutup ayısı öpermiş
    polar bear ****s unlucky nomad in the desert
    is it good?
    can google does it better

    ReplyDelete
  95. I think that all languages from all country in the world is very important and deserve to be added on google translate.

    ReplyDelete
  96. i think it shuould go from english to irish

    ReplyDelete
  97. It is great idea to add more languages.
    It will be good to add Macedonian language too

    ReplyDelete
  98. i like nooooooooodles

    ReplyDelete
  99. Gimme Malaysia language. It's not much different from indonesian language, so i dont think Google has to do a lot of works.

    ReplyDelete
  100. yaw zaten ingilizce bilsem turkce cevırıyı napayım..

    ReplyDelete