Google Translate Will Now Use Neural Learning to Make Translations Even Better
If you need to translate some text, Google Translate will usually help you roughly, but this is not ideal. The company now hopes to dramatically improve its translations using neural learning.
Google first announced that it was exploring ways to use neural networks to train its computers to translate skills back in September . Now the results of this program are being introduced into production. A new update to the Translator app will improve translations for and from the following languages:
- English
- French
- Deutsch
- Spanish language
- Portuguese
- Chinese
- Japanese
- Korean
- Turkish
This is a very small fraction of the languages that Translate can already work with, but it also accounts for a large percentage of the spoken languages on Earth. The goal of neural learning is to help the Translator understand the broader context of the translatable, rather than focusing on individual words or phrases and missing grammar.
Found in translation: More accurate, smoother sentences in Google Translate | Google via Android Police