Austronesian Language Translation [closed]

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance. Closed 10 years ago.

Google Translate is awesome for most of the major languages but where do you even begin if you wanted to make your OWN translation engine? Let's say I want to create a very basic Cuyonon to English or even an English to Cuyonon phrase translator,开发者_如何学C where do I begin?

---

You read about 5,000 pages on the science of machine translation. Google uses statistical machine translation. They collect gigantic parallel corpora of text in the two languages. They they match up the sentences (this alignment problem is not trivial) and then they train a gigantic statistical model. There are open source kits that can build these models if you have all the data, but they won't work as well as Google's.

For example, this.

---

I would petition google. If you are translating an aboriginal language you will require a character set. This may not exist. That is where I will start, unless anybody has a better idea? Following this I will make a database accessible from a web server (as I am a .NET developer I'm getting worried about the amount of apple handhelds used in the lands of the people this tool is for - another topic) and begin by doing a word for word translation of the most commonly used words by balander (derived unfortunately from Hollander - the name given to the people without souls). I agree with the first answer on required linguistical mechinisms, and would definately check open source modelling tools (or proprietry kits if it made sense). The amount of work required depends on who you know, and how you approach this - I'd be interested to hear how it goes - kudos!