diTii.com Digital News Hub

Sign up with your email address to be the first to know about latest news and more.

I agree to have my personal information transfered to MailChimp (more information)

Jun182018

Google Bring Offline Neural Machine Translation to Android & iOS

Offline translations are now a lot better with Google bringing neural machine translation (NMT) technology to on-devices, so you can get high-quality translations even when no internet connection is available.

This means, NMT will now run directly in the Google Translate apps on Android and iOS device. And, since each language set is just 35-45MB, offline translations are most useful when travelling to other countries without a local data plan, or when no internet or cellular data is available.

The neural machine translation significantly improved accuracy of online translations in Google Translate as it translates whole sentences at a time, rather than piece by piece. This makes translated paragraphs and articles a lot smoother and easier to read.

Here is an image of showing comparison between phrase-based machine translation (PBMT), new offline neural machine translation (NMT offline), and online neural machine translation (NMT):

Google Translate Comparison between phrase based and online-offline NMT

Google explaining the technology said, “It uses broader context to help determine the most relevant translation, which it then rearranges and adjusts to sound more like a real person speaking with proper grammar.”

To try out offline NMT translations, just update your Translate app on Android or iOS to the latest version. And, then head to offline translation settings and tap the arrow next to the language package to download it. Or, if you already have use offline translations earlier, a banner on home screen will take you to update your offline files.

In this screenshot you can see the banner notification on Google Translate home screen:

Google Translate offline NMT update banner

Google Translate offline NMT update is now available in 59 languages around the world.

Also, a recent Google publication revealing two new algorithms related to natural language processing (NLP) was published.

The first algorithm entitled, “Learning Semantic Textual Similarity from Conversations,” claims a new state of learning how to understand questions by studying responses as a way to understand what the questions really mean. Here’s how Google explains it:

“The intuition is that sentences are semantically similar if they have a similar distribution of responses. For example, “How old are you?” and “What is your age?” are both questions about age, which can be answered by similar responses such as “I am 20 years old”. In contrast, while “How are you?” and “How old are you?” contain almost identical words, they have very different meanings and lead to different responses.

You can read more about Google’s announcement on artificial intelligence (AI) on this Advances in Semantic Textual Similarity post.

Google Semantic Textual Similarity

In other news, Google added a new card in the “Family Link” app that shows parents “nutritious” apps—which are “recommended by teachers to help parents discover and determine if content is appropriate for their kids.”

This card is available now to all Family Link users in the U.S. with children ages six through nine. It’ll be rolling out soon to more countries.

An image of nutritious apps card in Family Link:

Google Nutritious Apps Card in Family Link App

Share This Story, Choose Your Platform!

Do NOT follow this link or you will be banned from the site!