Machine translation, sometimes referred to by the acronym MT, is a sub-field of computational linguistics that investigates the use of computer software to translate text or speech in between natural languages.
At its basic level, MT performs simple substitution of atomic words in one natural language for words in another. Using corpus techniques, more complex translations can be performed, allowing for better handling of differences in linguistic typology, phrase recognition, and translation of idioms, as well as the isolation of anomalies. However, current systems are unable to produce output of the same quality as a human translator, particularly where the text to be translated uses casual language.
Modern machine translation software, such as that produced by SYSTRAN or IBM, allows for customization by domain or profession (such as weather reports) — improving output by limiting the scope of allowable substitutions. This technique is particularly effective in domains where formal or formulaic language is used. Improved output quality can also be achieved by human intervention: for example, some systems are able to translate more accurately if the user has unambiguously identified which words in the text are names. With the assistance of these techniques, MT has proven useful as a tool to assist human translators, and in some cases can even produce output that can be used "as is".
In the words of the European Association for Machine Translation (EAMT):
The translation process, whether for translation, can be stated simply as:
Behind this simple procedure there lies a complex cognitive operation. For example, to decode the meaning of the source text in its entirety, the translator must interpret and analyse all the features of the text, a process which requires in-depth knowledge of both the grammar, semantics, syntax, idioms and the like of the source language, as well as the culture of its speakers. The translator needs the same in-depth knowledge to re-encode the meaning in the target language.
Therein lies the challenge in machine translation: how to program a computer to "understand" a text as a human being does and also to "create" a new text in the source language that "sounds" as if it has been written by a human.
This problem can be tackled in a number of ways.
Pyramid showing comparative depths of intermediary representation, interlingual machine translation at the peak, followed by transfer-based, then direct translation.
Machine translation can use a method based on linguistic rules, which means that words will be translated in a linguistic way — the most suitable (orally speaking) words of the target language will replace the ones in the source language.
It is often argued that the success of machine translation requires the problem of natural language understanding to be solved first.
Generally, rule-based methods parse a text, usually creating an intermediary, symbolic representation, from which the text in the target language is generated. According to the nature of the intermediary representation, an approach is described as interlingual machine translation or transfer-based machine translation. These methods require extensive lexicons with morphological, syntactic, and semantic information, and large sets of rules.
Given enough data, machine translation programs often work well enough for a native speaker of one language to get the approximate meaning of what is written by the other native speaker. The difficulty is getting enough data of the right kind to support the particular method. For example, the large multilingual corpus of data needed for statistical methods to work is not necessary for the grammar-based methods. But then, the grammar methods need a skilled linguist to carefully design the grammar that they use.
To translate between closely related languages, a technique referred to as shallow-transfer machine translation may be used.
Machine translation can use a method based on dictionary entries, which means that the words will be translated as a dictionary does — word by word, usually without much correlation of meaning between them.
Statistical machine translation tries to generate translations using statistical methods based on bilingual text corpora, such as the Canadian Hansard corpus, the English-French record of the Canadian parliament. Where such corpora are available, impressive results can be achieved translating texts of a similar kind, but such corpora are still very rare.
Example-based machine translation (EBMT) approach is often characterised by its use of a bilingual corpus as its main knowledge base, at run-time. It is essentially a translation by analogy and can be viewed as an implementation of case-based reasoning approach of machine learning.
Interlingual machine translation is one instance of rule-based machine translation approaches. According to this approach, the source language, ie. the text to be translated is transformed into an interlingual, ie. source/target language independent representation. The target language is then generated out of the interlingua.
The history of machine translation generally starts in the 1950s after the second world war. The Georgetown experiment in 1954 involved fully automatic translation of more than sixty Russian sentences into English. The experiment was a great success and ushered in an era of significant funding for machine translation research. The authors claimed that within three or five years, machine translation would be a solved problem.
However, the real progress was much slower, and after the ALPAC report in 1966, which found that the ten years long research had failed to fulfill the expectations, the funding was dramatically reduced. Starting in the late 1980s, as computational power increased and became less expensive, more interest began to be shown in statistical models for machine translation.
Today there are many software programs for translating natural language, several of them online, such as the SYSTRAN system which powers both Google translate and the AltaVista's Babelfish. Although there is no system that provides the holy-grail of "Fully automatic high quality translation" (FAHQT), many systems provide reasonable output.
Despite their inherent limitations, MT programs are currently used by various organizations around the world. Probably the largest institutional user is the >European Commission, which uses a highly customised version of the commercial MT system SYSTRAN to handle the automatic translation of a large volume of preliminary drafts of documents for internal use.
A Danish translation agency, Lingtech A/S , has been translating patent applications from English to Danish since 1993 using a proprietary rule-based machine translation system, PaTrans, working together with the translation memory based Trados commercial CAT tool. The system requires manual pre- and post-editing, but the monthly output is still approx. 400,000 words per operator.
The Spanish daily newspaper Periódico de Catalunya is translated from Spanish into Catalan with an MT system. .
Google has reported that promising results were obtained using proprietary statistic machine translation engine . However, it seems that the machine translation system they are currently using is still based on SYSTRAN engine.
It has been reported that in April 2003 Microsoft began using a hybrid MT system for the translation of a database of technical support documents from English to Spanish. The system was developed internally by Microsoft's Natural Language Research group. The group is currently testing an English-Japanese system as well as bringing English-French and English-German systems online. The latter two systems use a learned language generation component, whereas the first two have manually developed generation components. The systems were developed and trained using translation memory databases with over a million sentences each.
With the recent focus on terrorism, the military sources in US invest significant amounts of money in natural language engineering. In-Q-Tel  (a venture capital fund, largely funded by the US Intelligence Community, to stimulate new technologies through private sector entrepreneurs) brought up companies like Language Weaver. Information Processing Technology Office in DARPA hosts programs like TIDES and Babylon. US Air Force has awarded a $1 million contract to develop a language translation technology. 
Currently the military community is interested in translation and processing of languages like Arabic, Pashto, and Dari.
There are various methods for evaluating the performance of machine translation systems, the oldest is by using human judges to tell the quality of a translation, newer, automated methods include BLEU, NIST and METEOR.
Currently, the product of machine translation is sometimes called a "gisting translation" — unless one is proficient in both languages, MT will often produce only a rough translation that will at best allow the reader to "get the gist" of the source text, but is unlikely to convey a complete understanding of it. The user may find the raw translation sufficiently useful as it is.
From Wikipedia, the free encyclopedia