Mountain View, California -- Congratulations! Felicitaciones! Glückwünsche! -- Ever encountered a phrase in Esperanto, Bengali, Chinese, or some other language that needs translation? Well! Google's decade-long endeavor to automate the tricky business of translating languages appear to be paying off. The search engine behemoth over the weekend announced that its Google Translate service has reached an incredible milestone of more than 200 million active monthly users, and that mobile usage is quadrupling year over year.
Calculated roughly, the Web giant's translation service is deciphering more words every day that is enough to fill one million books than all human translators produce in a year.
“To put it another way: In a given day we translate roughly as much text as you would find in 1 million books. Besides, what all the professional human translators in the world produce in a year, our system translates in roughly a single day. By this estimate, most of the translation on the planet is now done by Google Translate,” writes Franz Och, Distinguished Research Scientist for Google Translate in a Thursday blog post.
The translation service initially unveiled in 2001 with just eight languages that could be translated to and from English. Back then the translations were of poor quality and anything other than single word and short phrase translation was a challenge. Writing on the official Google Blog, Och said this useful service has come a long way in the last 6 years, when the new translation engine was introduced.
Describing the service in detail, the post noted -- In 2001, Google started providing a service that could translate eight languages to and from English. It used what was the state-of-the-art commercial machine translation (MT), but the translation quality was not very good, and it did not improve much in those first few years. In 2003, a few Google engineers decided to ramp up the translation quality and tackle more languages.
Meanwhile, facing the challenge, Och had embarked onto retool the new system in the background, to improve the accuracy and usefulness of Google Translate, but this system was too slow to use for online services until 2006. In that year his new data-driven approach debuted in Chinese and Arabic language translations being offered online.
Och describes how slow his new system was before 2006: “…our system was too slow to run as a practical service--it took us 40 hours and 1,000 machines to translate 1,000 sentences. So we focused on speed, and a year later our system could translate a sentence in under a second, and with better quality. In early 2006, we rolled out our first languages: Chinese, then Arabic.”
After diligently working on improving speed and the accuracy of its translation machines, Google bundled more languages. Today, the service can translate a sentence in less than one second and works in 64 different languages, including Icelandic, Swahili, Basque, Azerbaijani, and Welsh.
Apart from constantly enhancing the service, “We rolled out our statistical MT approach on April 28, 2006, and in the six years since then we have focused mainly on core translation quality and language coverage. We can now translate among any of 64 different languages, including many with a small web presence, such as Bengali, Basque, Swahili, Yiddish, even Esperanto.”
According to Och, he considers that the majority of Google Translate users access the service while traveling. And as per the company's stats, mobile traffic has quadrupled during the past couple of years and 92 percent of traffic comes from outside the U.S.
Besides, Google Translate is available for texting on mobile phones, in Google Chrome, the browser in ascendance for any Web page, YouTube video captions, and speech-to-speech “conversation mode” on smartphones.
Nevertheless, Google is continuing to dream of a society where it does not matter what languages a group of people speak, they will all be able to understand and contribute to a conversation, using a form of a translation program. Preferably Google Translate, of course.