Machines or Translators


google-vs-bingThe EU Institutions require strict precision in their documents and translations. To ensure this, their translators use particular software, applications, databases and glossaries, each language is dealt with in a separate language unit, and translators have busy, frenzied days of work. But what about the web community, researching, communicating and reading papers in different languages? Most of the users are quite satisfied with online tools, that permit a fair, even though not perfect, translation of a common document.

By now, two are the main competitors in the battleground of online translations. The usual Google, with its huge amount of stored countless archives and data, and Microsoft’s Bing, that made a successful and strategical deal with Facebook to be the chosen tool of choice in Zuckerberg’s network, in order to duel with the giant of MountainView, which is still the tool of choice for millions of users. Despite the software, anyway, both use, for the most, what is called “statistical machine translation”. Google, particulary, here has a lion’s share. With an impressive velocity and advanced algorithms, the search engine “looks for patterns in hundreds of millions of documents to help decide on the best translation for you. By detecting patterns in documents that have already been translated by human translators, Google Translate can make intelligent guesses as to what an appropriate translation should be“. This is also a quick, simple definition of statistical machine translation. If you think that mechanical pattern recognition in some way exploits the labour of hundreds of translators, well, maybe you are right. Still, it’s something that can work and help millions of simple users to surpass cultural and linguistic barriers.

Literally, millions can have conversations and exchange documents without boundaries. Currently Google can translate documents in 80 languages, from Hausa Nigerian dialect (35 million native speakers) to Punjabi language (100 million speakers in India and Pakistan) to Maori (160 thousand), among others. This is made possible thanks to Wordnet licences and open source components. Facebook, on the other hand, connects almost 900 million active users, that can count on a translation by Bing in the case of a foreign language. Big numbers.

And more is yet to come. Documents and translations are constantly uploaded, with an amazing growth, that Google itself can analyse only retrospectively. Moreover, one of the instruments for perfecting translation is the direct calls to translators and the users themselves, that can modify, correct and record a new improved translation. Even if the human touch is still necessary, and still will be in crucial official and legal documents, the future of translation seems to be more and more technology related.

As suggested by one of our terminologist bloggers, Licia Corbolante, look at this visual explanation, Inside Google Translator.

more on google translator

By Matteo Poles

Social Media Specialist

Communication Trainee at TermCoord