Facebook IT & Web-tech News Research

Facebook translation tool is using machine learning to serve 2 billion text translations per day

facebook
Share on Facebook
Tweet about this on TwitterShare on Google+Share on StumbleUponShare on LinkedInPin on PinterestShare on Reddit

Facebook aims to connect the whole world through its social networking platform. Towards that objective, to solve one major hurdle of the language barrier, Facebook is using machine learning based translation tool to aid translations on its network. And the response has been no less than overwhelming as 800 million users are already using the feature with over 2 billion text translations per day.

This information  was recently revealed by Alan Packer, Facebook’s director of engineering for language technology, during MIT’s EmTech Digital conference in San Francisco.

Packer said that previously Facebook was using Bing’s translation tool but found out that the tool was too focussed on website based formal language and was not suitable for casual language which people usually use in their day to day conversations.

According to Packer,

“Bing didn’t do well on slang, idioms, and metaphors. We really needed to train on our own data.”

With about two trillion posts and comments on Facebook and with a billion more being added each day by 1.6 billion users, Facebook already has access to tons of user data on its platform. It began feeding this data to its machine learning system to train it for translations.

Last year, to boost its efforts in this direction, the company also acquired a startup called Wit.ai which was using the understanding of natural language in text and voice to power new user interfaces.

As a result of this, Facebook was able to develop its own translation tool which is able to translate text based on context and relevancy of conversation. It began using the tool from December last year and now it is fully rolled out for 1800 different translation permutations.

The Facebook translation tool can understand more than 400 languages and dialects and translate across 40 languages in 1,800 different directions.

Half of the total users of Facebook (1.6 billion) are already using the tool which now serves 2 billion translations in a single day. Users can access the tool from the drop down menu over any post in any foreign language and click on “See Translation” button to see the post in their own native language.

The objective behind bringing such feature to its network is clear- and that is to make the world more open and connected.

“The mission of the translation team is removing language as a barrier to making the world more open and connected,” said Pecker.

He further added that users are able to “have more friends, more friends of friends, and get exposed to more concepts and cultures” using such translation product.

[email protected]


Add Comment

Click here to post a comment

Your email address will not be published. Required fields are marked *