Experts at Facebook opted mathematic algorithms for better translation

The world is a huge place with a lot of inhabitants living on it. Each region has its own language. Some are popular and others don’t even diffuse from their own habitat. So, they are not even included in the modern dictionary. That is the reason, they are not being translated.

Mathematical algorithm for better translation:

Facebook is indulged in finding an alternative way for translation of those languages, which are not even mentioned in modern dictionary.  Facebook already has 200 languages on the board but that’s not enough. The automatic translation is used now a days and it is based on having large storage of similar words from different languages. Sometimes, there comes a language which does not have any identical words.

Therefore, in order to solve this problem researches at Facebook are being conducted to find an alternative way for translation. They have sought to use mathematic algorithm for solving this problem. According to the researchers, every word will become a vector quantity in a space which will be multidimensional. Now, it will work based on proximity. The words or vectors which are similar to each other will lie at close proximity in multi-dimensional space. And when the two vectors come to a common space, it means they are the same words from different languages. This is a bit difficult task for the researchers to make such an algorithm and multidimensional space, but it’s not impossible.

must see? Facebook: A Milestone Of 5 Billion Downloads on Android

Many popular firms like Facebook, Google, Microsoft, Yandex, Baidu are constantly involved in making their translation tools better. So, Facebook has provided them with a very bright and applicable idea. The experts around the world are convinced that this word vector technique can bore fruit. It can enable us to understand the languages of the most remote areas.

Leave a Reply

Your email address will not be published. Required fields are marked *

Close

Adblock Detected

Please consider supporting us by disabling your ad blocker