Facebook’s Translation Tool: Review

In recent news, Facebook has updated their site to include a translation service to translate posts from one language to another. Personally, I have been looking forward to the introduction of this service as I have several Facebook friends for whom English is not their first language. While these friends are happy to communicate with me in English, I do miss out on being a part of their day to day status updates in their native language.

To enable this feature, log in to your Facebook account and then click on ‘Your Settings’, and navigate to the ‘Allow translations from’ options. Then simply choose from any or all of the following options: ‘Admin’; ‘Community’; or ‘Machine Translators.’
 
It seems fitting that Facebook is the first to roll out an extensive and relatively sophisticated translation service for social media. After all, Facebook was the social media company who first allowed everyday people to communicate with others all over the world, so it is a logical next step to then allow this communication to cross language divides.
 
In the minds of many avid Facebookers, this feature was long overdue. But how does it work behind the scenes, and has it been thoroughly thought through?
 
Facebook has teamed up with Bing to provide this service. All of the translations are handled by Bing, though Facebook users can edit translations and even add their own. The idea is that bilingual Facebook users will step up and change translations to make them as accurate as possible so that people who are unfamiliar with a certain language will get the most accurate translation. This kind of crowdsourcing can be both exciting and fraught with trouble.
 
The danger is that erroneous translations, whether purposeful or accidental, will make it past Bing’s filter, giving unsuspecting users the wrong translation. Bing has tried to counter this possibility by using a voting system whereby users vote for the most accurate translation for each phrase that has been entered. The theory behind this is that erroneous translations will not be voted as the most accurate translation, ensuring that only the best and most correct translation will rise to the top of the list.
 
However in my opinion Bing’s fallacy lies in the fact that the overwhelming majority of people who vote for translations will not be truly bilingual. It is recognised that only bilingual people will be able to accurately gauge whether a translation is correct, but they will make up only a tiny proportion of the voting population.
 
The result? People who are all but ignorant of one of the languages involved in the translation may be determining the ‘best’ translations. While the potential exists for bad translations to become popular, I think that the more pressing issue is the problem of the best translations not becoming the ones used in Bing’s system.
 
My advice is that Facebook’s new translation service is the perfect tool for you to share your personal Facebook status updates with your multilingual friends, and for you to become a part of your friends’ other-language online lives. However, I would be wary of allowing this new tool to go anywhere near your business’s status updates, or anything that impacts your professionally. Crowdsourcing is great for what it is, but in my opinion it should be confined to personal use where the odd error will not impact on your delicate professional reputation.