On Jan 8, 6:50 pm, "goncado...@gmail.com" <goncado...@gmail.com> wrote: > I think that would be great if google translate will consider to add > Turkish on language scala. More than 80 million people speak turkish, > much more than the some other languages. This tool can be really > helpful for turkish people and needed. > > What is the process of adding new languages?
The first thing they need is a large corpus of bilingual text. They got this from UN and EU for many countries, but also probably from other sources - some nations, such as France, use a lot of resources on documenting their language out of nationalism (and because they are pretty rich), while others, such as the Philippines, have had to do a lot of public language work out of necessity even though they are poor. This is the only way we can help them, if we have a private database of bilingual text we can give them access to. Then they need a huge body of text in the new language. I don't believe they have any problem of getting this for Turkish, they're Google after all. Then they need to adapt their system to the quirks of the new language. From what they have said publicly, I believe Google's translator operates on overlapping sequences of eight words. This will create trouble in languages where the verb comes on the end of the sentence, and the sentence is longer than 8 words. I know German has this problem, and that Google translate has found some kind of workaround - but maybe that workaround isn't so easy to transfer to Turkish. Also, some languages tend to glue together words to form more complex concepts. Turkish is famous for it. This would lead to the translator frequently running into words it has never seen before. Probably they deal with this by custom rules for "picking apart" compound words. These would be specific to each language. Maybe they haven't made a good one for Turkish yet. But Turkish will come, eventually, I'm sure of it. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Google Translate" group. To post to this group, send email to google-translate@googlegroups.com To unsubscribe from this group, send email to google-translate+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/google-translate?hl=en -~----------~----~----~----~------~----~------~--~---
