News

Google Translate Gets Smarter; Integrates Neural Machine Translation To Interpret Local Slangs, Vernacular

Google Translate Gets Smarter; Integrates Neural Machine Translation To Interpret Local Slangs, Vernacular

In communication, deciphering a speaker's tone is subjective. Google understands that no two vernaculars reflect the distinctions of either, hence, the re-birth of Google Translate.

In the past,, Google Translate simply tried to decipher chunks of text, producig quite comical results. The translated sentences are not constantly precise but are handy enough to be comprehensible for native tongues. The old algorithm tends to disregard the compound phrasing, context and syntax of a sentence. As a result, so many specifics got lost in translation.

Google Translate No Longer Lost In Translation

Google launched the Neural Machine Translation feature to try to improve on its original application. Aside from the old phrase-based Machine Translation in the application, Google will now integrate the Neural Machine Translation. This will employ pioneering training methods to convey enhanced machine translation.

Basically, Neural Machine Translation will try to translate an entire sentence at a time, rather than the old word for word translation technique. Google Translate, with this upgrade, will now use larger context to define the most relevant translation there is. It arranges and fine-tunes the translation and bundle it with appropriate grammar to sound more natural and human. Users can expect that as Google Translate is used frequently, the machine gets better in its translating capability.

Google Translate Neural Machine Translation Now In Eight Languages

Barak Turovsky, product lead of Google Translate, said that the Neural Machine Translation is now available in eight languages. People can now translate from English and German, French, Spanish, Chinese, Portuguese, Korean, Turkish and Japanese.

The available languages comprise 35 percent of the entire translations done on the app. Google wants to eventually bring the technology the 103 languages accessible to users. The company is also introducing its machine learning technology through the Google Cloud platform. Through this, developers will be able to fiddle their Machine Learning APIs.

© 2024 Game & Guide All rights reserved. Do not reproduce without permission.
Join the Discussion
More Stories
Real Time Analytics