What is the mechanism of Google Translate service?
Incredible machine learning technology behind google translator.
It is not possible that they have not used even once, at the very least your whole life metropolitan By using the Internet, the site that does not disappear for a moment from the date browse any personal computer or virtually mobile phone, do not talk today about the Google search engine itself, but we are talking about a The most important services provided by this great edifice which, a simultaneous translation service, known as Google Translate. Certainly, we are not here today because we definition service, as almost already said not without a list browsing history of this site, but I today I will try to answer the question you may have the risk to your mind days which is: What are the mechanism of action Google Translate service and similar services?
Could you think of first glance that there is a place where to meet thousands of employees from all over the world and speak all languages, words are added daily to the site by hand so that the location of the translation of each word asking him to translate? Well, if we talked about in your imagination separate, yes, there are thousands of employees from all over the world are working with Google, but certainly not their job is to enter data by hand, what you see is how they carried out the site to translate every word is entered to him? Next lines to learn.
About the service and its history:
Google Translate is a free service developed by Google, you can translate the texts and conversations, images and locations in real time, between 103 supported languages in addition to the fourteen languages are developed. The service was first launched in April 2006, meaning that in April of the previous year was the tenth anniversary of its creation. Visit the site and is used by more than 200 million visitors a day. Service initially started as a Statistical Machine Translation, turned up in November 2016 to Neural Machine Translation.
Statistical machine translation - Statistical Machine Translation:
The beginning of this technique at the beginning of the year 2000, where the translator machines to develop complex algorithms for more accurate translations, but it is, of course, did not live up to the translation, such as the sons of man. The site works like the rest of the translation sites through statistics. And through the introduction of millions of documents and books that have been translated by humans. In addition to United Nations documents, which are translated into six major languages, and then providing the linguistic data. And thus able to provide the location of many of the translation options for the sentences, which have been submitted already - the Pre by humans, and then choose the closest these translations and the most commonly used and common.
Here lies the difficulty, the matter is not as I have already mentioned an alarming number of people are sitting to enter words and their meanings, word for word, but the difficulty lies in finding this huge linguistic content Therefore, it has been the use of English as a mediator between any two languages, so that if you want to word translation from Arabic into French, for example, will be Arabic translation first into English, and then will be translation from English to French, and, of course. Add to that, that there is some language in which the number of operations for two more. For example, if you want to translate the word of the Catalan language to Chinese, will be the site first translation of the Catalan word to the Spanish word, which is the origin of this language in this case, it then translates the Spanish word to the English word, and then from English to Chinese, and it is also in a fraction of a second.
google translator neural network
how translator works
Mahmoud Emad Elerian
writer at newsandstory.com
Total 6730 views