Google has introduced new neural network machine translation technology to translations between English and nine widely used Indian languages – Hindi, Bengali, Marathi, Gujarati, Punjabi, Tamil, Telugu, Malayalam and Kannada.Neural machine translation translates full sentences at a time, instead of pieces of a sentence, using this broader context to help it figure out the most relevant translation. The result is higher-quality, more human sounding translations.
The company has also extended its new translation technique’s support to Chrome browser’s built-in auto-translate functionality. Additionally, it has announced that its Gboard Keyboard app will now support all 22 scheduled Indian languages. Finally, Google Search results will now include results from Hindi dictionary as well.
More than 400 million people in India use the internet, and more are coming online every day. But the vast majority of India’s and global online content is in English, which only 20 percent of the country’s population speaks—meaning most Indians have a hard time finding content and services in their language. With the translation support Googe aims to make english content more accessible to the new users and increase its user base.
Just like it’s easier to learn a language when you already know a related language, our neural network technology speaks each language better when it learns several at a time. For example, we have a whole lot more sample data for Hindi than its relatives Marathi and Bengali, but when we train them all together, the translations for all improve more than if we’d trained each individually.
Neural networks approach the problem in a different way. The idea is to take a large number of data set, known as training examples,and then develop a system which can learn from those training examples. In other words, the neural network uses the examples to automatically infer rules for recognizing the data.