Earlier this year, Facebook launched a “multilingual composer” to let Pages share posts with diverse audiences in many languages, automatically serving individual users the post in their preferred language.
Starting on Friday, millions of Facebook users who aren’t Page holders will also be able to use the multilingual composer by turning on the feature in their language settings, the Menlo Park, Calif.-based company said in a post. The social media giant decided to offer the composer more widely to cater to users with close friends and family who speak multiple languages. The tool makes it possible, for example, for a multilingual post created in English and Spanish to be delivered to English speaking users in English and to Spanish speakers in Spanish. Facebook uses signals such as users’ preferences, location and the language used in their posts to determine their primary tongue. Facebook plans to gather more feedback from users before setting a timeline for global rollout of the tool.
“Given that 50% of our community does not speak English and most people don’t speak each other’s languages, it’s incredibly important that we take on the challenge of removing language as a barrier to connecting on Facebook,” the company said. ”We’re excited to see this tool help even more people connect with their friends who speak different languages.”
The composer uses machine translation to pre-fill messages in additional languages selected by the user and taps into the same system that generates “See Translation” results across Facebook. Before Facebook offered the composer, Pages would often write posts in multiple languages and use targeting to direct posts at specific audiences. Pages would also publish long posts containing the same message with many translations. The composer tool has eliminated this work for many public figures and organizations, and is now used by about 5,000 Pages to post an average of 10,000 times daily.
Facebook said it plans to use multilingual posts to improve its machine translation data.
“We hope to collect new data in less common languages, enabling us to build new machine translation systems that work better,” the company said. “This will move us closer to our vision of removing language barriers across Facebook.”