Littera Deusto

Modern Languages, Basque Studies and Humanities

Questionnaire 2: Machine Translation (MT)

marzo 15th, 2009 · No hay Comentarios

Now, we have a clear idea of what Natural Language Processing (NLP) is; besides, we have already mentioned the most common tasks that have to do with NLP. In this post, what I would like to do is to start explaining briefly some of those topics for you to have a better understanding of them.

Trans·la·tion

Trans·la·tion

Machine Translation (MT):

Machine Translation (MT) is a sub-field of Computational Linguistics whose main purpose is to develop computer software in order to translate automatically text or speech from one natural language to another one. At a first approach, MT substitutes words in one natural language for words in the target language. However, with the use of several corpus techniques we can obtain more complex translations.

The translation process could be stated as a two-steps operation. Firstly, the translator must decode the meaning of the source text; secondly, it must re-encode the original meaning in the target language. Explained in this simple way it may sound as a simple process, but in order to achieve successfully this task the translator must be able to interpret and analyse carefully the whole text; for this reason the machine needs to have in-depth knowledge of the grammar, semantics, syntax, idioms… of both the original and the target languages.

MT software has improved dramatically in the last years and, nowadays, it enables users for customization by profession, for instance, which makes the scope of allowable substitutions more limited. Consequently, we are progressively obtaining better results each time. However, MT faces two major problems: word sense disambiguation and named entities recognition. However, I will talk about them soon.

Summing up, and as I have mentioned above, MT consists of translating texts or speech from one natural language to another one using translation software. I wanted to present this topic here becuase I find it fascinating and I think that it will gain much more importance in an early future. And, of course, if you want to know more about it feel free to investigate. You will not regret!

References:

(Note: in the future this article could be modified several times in order to include more up-to-date information.)


Etiquetas:

  • Etiquetas