Facebook Creates Faster And Even More Accurate Translation System And Guess What It Uses!

Facebook Creates Faster And Even More Accurate Translation System And Guess What It Uses!

Facebook has more than a billion users and it is no surprise that they speak a multitude of languages. In such cases, it is important for Facebook to offer support for multiple languages lest its popularity is bound to suffer.

Currently, Facebook supports as many as 45 different languages and if you come across a post in a different language, you will always see a link that reads, “See translation.”

It was on this Tuesday that Facebook abounded that they have revamped the translation system and made it a whole letter better, faster and much more accurate as well. They have made smart use of artificial intelligence techniques to come up with the new, advanced and better translation system.

The FAIR group which actually means Facebook AI research has been monumental in this new translation system. Michael Auli, the key researcher working at FAIR was quoted as saying, “Neural networks are modeled after the human brain,”

This network can solve a multitude of tasks and one of these is the translation. They are also capable of solving different problems, summarizing texts and do a whole lot more. However, there are a lot of different kinds of neural networks. The central approach so far has been using recurrent neural networks which help in translating text by taking one word at a time. However, now the researchers at Facebook have come up with a different technique which is known as the convolution neural network. These look at words in the whole group rather than concentrating on it one word are time.

The researcher said that these networks look at the whole data at the same time and then they go for translation. This is bound to bring in quicker and more accurate translation which is sure to be applauded.

Facebook hasn’t yet said as to how it plans to integrate this new technology with the final product, but they have released the tech as open seduce so that all other coders too can make the most out of it.

Share this post

Leave a Reply