Thursday June 02, 2005

Coming soon : (improved) Google Translator

slashdot.org -> blog.outer-court.com :

At the recent web cast of the Google Factory Tour, researcher Franz Och presented the current state of the Google Machine Translation Systems. He compared translations of the current Google translator, and the status quo of the Google Research Lab's activities. The results were highly impressive. A sentence in Arabic which is now being translated to a nonsensical "Alpine white new presence tape registered for coffee confirms Laden" is now in the Research Labs being translated to "The White House Confirmed the Existence of a New Bin Laden Tape." ...

How do they do that? ... to the translation system, any language is treated the same, and there is no manually created rule-set of grammar, metaphors and such. Instead, the system is learning from existing human translations. Google relies on a large corpus of texts which are available in multiple languages.

This is the Rosetta Stone approach of translation ... all it needs is someone to feed the system the two books and to teach it the two are translations from language A to language B, and the translator can create what Franz Och called a "language model." I suspect it's crucial that the body of text is immensely large, or else the system in its task of translating would stumble upon too many unlearned phrases. Google used the United Nations Documents to train their machine, and all in all fed 200 billion words. This is brute force AI, if you want – it works on statistical learning theory only and has not much real "understanding" of anything but patterns ...

- So, when will this be rolled out? Will the biblical Tower of Babel thing happen again or is that passé already, with the successful completion of Taiwan's Taipei 101 ... :)

(2005-06-02 13:25:35 SGT) [Tech] Permalink

Most popular blog postings on lowem.log :

1. Singapore MRT rail network length to double by 2020
2. 2010 Nissan Leaf electric car specifications : 107hp, 24KWh lithium-ion batteries, 100-mile range
3. Live spot gold price quotes chart on COMEX
4. 2010 Toyota Prius specifications released : 50 mpg, 1.8L, 134hp, Ni-MH, solar roof option
5. AVG Anti-Virus Free Edition 2011 direct download link
6. Real-time live gold and silver price quotes chart on COMEX
7. Singapore electric vehicles : Government agencies EMA and LTA to study EV introduction
8. Book review : Shut Down by William Flynn

Featured articles on lowem.log :

1. Book review : Shut Down by William Flynn
2. Singapore electric cars testing starts with 9 electric vehicles
3. Honda, GS Yuasa JV to make lithium-ion batteries for 2010/2011 Honda Civic Hybrid
4. 2010 Honda Civic Hybrid preliminary specifications released
5. 2010 Honda CR-Z hybrid, 2010 Honda Fit/Jazz hybrid models confirmed
6. 2010 Toyota Prius specifications released : 50 mpg, 1.8L, 134hp, Ni-MH, solar roof option
7. NYMEX crude oil recovers from $32.40 low after 2.2 mbpd OPEC production cut announced
8. Singapore : Nuclear power not ruled out