“Facebook was not originally created to be a company. It was built to accomplish a social mission – to make the world more open and connected.” – Mark Zuckerberg – Co-founder & CEO – Facebook.
Facebook is the most popular social network in the world. On September 30, 2017, Facebook announces 2.07 billion active users per month. This is a great achievement, because no other company can now dream of such a base, that is so huge and yet so essential for all of them. They are the only ones responsible for changing the way we communicate with our loved ones. Facebook has also changed the definition of "friend" as "friend", and despite the use of "like" and "share" in the same sense, Facebook weighed completely.
As the Facebook community has grown decisively over the years, you must face some posts in your Facebook life that is in foreign language and looks strange to you? But what do you do when this publication appears in your news source? Simply click on a button and translate it in your own language. However, in addition to another host of online translation services, the Facebook translation service included an alert: it was not always logical. After all, a precise and accurate translation of about two billion contacts is a hard cut to break, especially when it comes to calculating goals, abbreviations, errors, spoken language and language at the same time. However, when Facebook decided to update its game and the power of neural networks, it may have translated it to a large extent.
The following is the translation power for the Facebook audience: Italy, on November 13, 2017, Sweden scored 0-0, for the first time in 60 years, it is enough for the 2018 World Cup. This has generated negative comments from many people around the world in social networks. Facebook and Twitter have been fooled by the wrath of fans who have asked for big names who have won the FIFA World Cup. Fans from around the world can exchange views on this topic and, according to the Facebook translation, they do not know all the language barriers.
The translation tool on Facebook, as you know, was launched in 2011, and Microsoft, Microsoft's affiliate, was at that time responsible for the translation. Initially, Facebook users can use the translation button along with comments written anything other than the English language simply, and translations appear in a separate window, so that you may compare it with the original text. Recently, Facebook has more than 4,500 million automatic translations and uses smart networks to perform this massive work with precision.
Just like during the start of this year, somewhere in January, Mark Zuckerberg gave out a bit about what might happen and coming next in Facebook, he clearly stated during some press conference, “Today we’re publishing research on how AI can deliver better language translations. With a new neural network, our AI research team was able to translate more accurately between languages, while also being nine times faster than current methods.”
What is surprising about this feature and how does it translate into text translation on the Internet? After all, we can use Google Translate and other free services. Facebook is very confident in this system, because it is believed that this new tool can change the way you use Internet translation services. So, from the automated machine translation model, based on the simple term, Facebook's focus is now based on neural networks. Let me set this point now so that everyone can understand why there is a lot of hope for Facebook in the future.
Facebook has truly incorporated neural networks in its translation tool. It is a difficult system of mathematics that is usually capable of learning things once it analyze some massive amount of data. So, if you are hiding somewhere behind the door and still oblivious towards this earth shattering discovery, let me assure you, this strategy is already in use in multiple fields and has helped in embracing the image and speed recognition, only for the better. Just like that, it is now used by Facebook for massive machine translations.
Google also tried to benefit from neural networks in 2016 and launched a new translation system based on this popular innovation. Google successfully completed the current forms at that time, such as those used by Microsoft. Facebook is trying to raise its profile and use neural networks to integrate artificial intelligence into its translation services. This is what you try to calculate errors, spoken language, text, etc. LSTM is what perfectly translates the Facebook site.
The attention span is the major distinction between the new and the old system. All the phrase based translations are now being replaced by the ultimate neural networks that can translate the whole paragraphs and take much more than few words at once. In this new way, some sort of LSTM, a kind of machine learning component, is being applied now. The advantages of it are quite clear as we get a whole accurate sentence after translation, getting the full context of the original one.
Facebook explained few factors of this network interestingly that, when the word that is written does not have a direct translation in a target language you are looking for, the neural system will then place any unknown word for it. The translation are commonly taken from the in-house dictionaries that is built from Facebook’s training data, and then the unknown word gets replaced.
Neural networks is the opportunity for many developments and is a growing area, which is related to even further context, combing the text of a post, in order to create some better translations.
People around the globe communicate with each other in many different languages, it is important to understand multiple ideas and choices, Facebook understands that and has thus bought this amazing change.