No two languages mirror the nuances of either, and Google knows that to be true. We've all, at some point in the past, attempted to translate blocks of text using Google Translate only to end up with amusing results.
Often, the translate portion really does simply "translate" the text, with the algorithm prone to disregarding the complex phraseology, syntax and context of a sentence, giving a jagged, corrosive and half-polished conversion as a result. Needless to say, so much details are lost in translation and sometimes even errors get in the mix of things — further obscuring the intended message.
Luckily, Google is all too familiar with this problem and is introducing Neural Machine Translation to try and enhance its pioneering application, Google Translate, which has soared to over 103 languages in just a decade.
Neural Machine Translation
At present, Google Translate uses Phrase-Based Machine Translation which aims to reduce the language barrier carried by longer sentences and complex contextual insertions in a given language. While this worked to a certain degree of acceptability, it was never a perfect and error-free method of translation.
Simply put, Neural Machine Translation tries to translate whole sentences at a time instead of a word-for-word translation method, which often leads to errors. It uses a broader context to determine what the most relevant translation is. It doesn't stop here, however. It then arranges and adjusts the translation to package it with proper grammar as to sound more human-like.
Best of all, the Neural Machine Translation gets better at translating sentences as people use it more frequently, because end-to-end to learning system is built onto Neural Machine Translation.
Availability Of Neural Machine Translation
"Today we're putting Neural Machine Translation into action with a total of eight languages to and from English and French, German, Spanish, Portuguese, Chinese, Japanese, Korean and Turkish," wrote Barak Turovsky, product lead of Google Translate, adding that these languages represent 35 percent of all translations made on the app.
Google eventually wants to bring the technology to all of 103 languages available for people to access. Google is also opening up its machine learning technology via the Google Cloud platform which offers Machine Learning APIs for developers to tinker around with.
Do you think Google Translate can solve language barriers in the future? Feel free to share us your thoughts!