Google: AI helps Google Translate offer these new languages spoken by millions

It's the first time in four years that Google has added Google Translate support for new languages.

For the first time in four years, Google is launching five additional languages for Google Translate, bringing its total number of supported languages to 108. 

Google says the large interval since it last added new languages is down to how its machine-translation app learns new languages. If there's a shortage of content on the web from a specific language, Google Translate doesn't have the data to build support for it. 

However, Google argues that recent advances in machine learning and help from members of the Google Translate Community have enabled it to launch the five new languages, which are Kinyarwanda, Odia, Tatar, Turkmen, and Uyghur. 

The languages are spoken by around 75 million people. Kinyarwanda is an official language of Rwanda that's spoken by about 12 million. Odia is a language spoken by 40 million in the Indian state of Odisha. Tatars, a Turkic language, is spoken by some seven million people in the Tatarstan region of Russia. 

SEE: How to implement AI and machine learning (ZDNet special report) | Download the report as a PDF (TechRepublic)

Turkmen is the official language of Turkmenistan, which is home to most of the world's seven million Turkmen speakers. Uyghur, another language from the Turkic family and an official language of Xinjiang Uyghur Autonomous Region of Western China, is spoken by about 15 million people. 

Google Translate supports text and website translation for each of the five languages. It also supports virtual keyboard input on a device's screen for Kinyarwanda, Tatar, and Uyghur, which join over 70 other languages that have on-screen keyboard support. 

Google will launch the new Translate languages for 1% of users today and then ramp it up to complete coverage soon after. 

The new languages widen the language support gap between Google Translate and Microsoft Translator, which supports over 60 languages. 

Microsoft in January added Irish Gaelic to Microsoft Translator, allowing Microsoft customers to build translation systems to their products. It too attributed new language support to recent gains in neural machine-translation technology, which allowed to it to use deep learning to improve its machine-translation models.      

While end-users gain the ability to translate text in other languages for free, both companies are offering translation platforms to support paid-for products that help businesses reach customers across languages. 

SEE: Google: Translate will soon give you near real-time transcriptions

Microsoft has the Translator Text and Speech API that are part of its Azure Cognitive Services stable. Google offers the Translate API and AutoML Translation. The AutoML Translation product caters to businesses with little machine-learning expertise, enabling them to build custom translation models. 

Amazon Web Services also recently added support for 22 new languages to its real-time translation service Amazon Translate, bringing the total number of languages it supports to 54 languages and dialects. Industrial manufacturing giant Siemens uses the AWS service to analyze employee surveys in different languages.