Machine translation (or MT) is translation of text that is processed completely by computer software, from its original language to its target language. There is no human intervention in MT. It is also known as automated/ automatic translation or instant translation. It has been developing since the 1940’s, and has been evolving steadily with new techniques, capabilities and technical advancements in its field.
The different approaches to Machine Translation
- Rule Based Machine Translation (RBMT)
- Statistical Machine Translation (SMT)
- Syntax Based Machine Translation (SBMT)
Out of the 3 different approaches SMT has proved to be the most popular and has been refined the most, but it still has its challenges; such as the terminology and writing styles are inconsistent and unpredictable.
There are certain benefits with MT; the most significant being speed. The amount of text that can be translated by a human in one day would be the amount a computer can in one minute. In the case of high quality machine translation process (which would include post editing) the accuracy and quality of the translation is considered much higher than pure human translation. MT also helps lower translation costs.