Facebook using neural networks to power translation systems
Facebook has switched from using phrase-based machine translation models to using neural networks to power its translation systems. In the past, Facebook used machine translation to translate text in posts and comments automatically, breaking language barriers and allowing people around the world to communicate with each other. However, creating seamless, highly-accurate translation experiences for the 2 billion people who use Facebook is difficult, and it needed to improve its systems.
“We need to account for context, slang, typos, abbreviations, and intent simultaneously,” Facebook said. “To continue improving the quality of our translations, we recently switched from using phrase-based machine translation models to neural networks.” Their platform now accounts for more than 2,000 translation directions and 4.5 billion translations each day.
“These new models provide more accurate and fluent translations, improving people’s experience consuming Facebook content that is not written in their preferred language,” the company said.
- ^ Facebook has switched (code.facebook.com)
- ^ Facebook launches new algorithm to fight fake news (mybroadband.co.za)