Machine translation is a subfield of Computational Linguistics, an interdisciplinary field combining mainly computer science and linguistics (you can read more about it here). There are various different approaches to machine translation, including the rule-based machine translation, statistical machine translation (which are both explained in the lesson), as well as example-based machine translation and hybrid machine translation.

The most important problem for a machine translator is trying to accommodate for all the irregularities and exceptions to the rules. If a language did not have any of these, it would be a piece of cake to build a machine translator! Human languages are living things, meaning they are constantly changing, because they are used in everyday life. New words, phrases and expressions are coined and old ones fall out of use. Thus, it is difficult to build a machine that accounts for all these irregularities, exceptions and changes. Read this article to find out some of the problems with machine translators!

Several machine translators are out there, free for use. The most popular one is perhaps Google Translate, but there are others like SYSTRAN. Give them a try and see how accurate they are! Know a second language? See if the machine translator can correctly translate between your native language and your second one. What was the result? Would you trust it to do your homework or talk to a friend from another country for you? Will machine translators bridge the barrier between languages especially online?

Next Section »