Machine Translation (MT) has been around for some time now. The first experiment with Machine Translation started in 1954. The IBM – Georgetown experiment involved the 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 in the United States.
The US Scientists claimed that within 3-5 years machine translation would be a solved problem. Around the same time, work started in Moscow on similar projects. The emphasis in both countries was the translation of Russian-English pairs. It is a well-known fact that war fare and defence drives technology and this was certainly the case in getting MT off the ground.
Real progress, however, was much slower. In the 1966 the ALPAC report found that 10 years of research had failed to fulfil the early promise. The increase in computing power in the 1980’s generated more interest in Statistical MT.
Moving forward another 30 years, we still don’t have a definitive MT tool but there are many programs now available that are capable of providing useful output within strict constraints; several of them are available online, such as Google Translate and BableFish.
As we can see, MT has developed from a tool used by Cold War spies into a widely used tool available to everyone. A back-packer in the outer reaches of Vietnam can now type a phrase into Google Translate and ask for a beer and a meal in Vietnamese. It may not be totally correct but with the ability to hear the phrase spoken the boundaries of language between people are disappearing. This can only be a good thing.
As a translation agency embrace this new technology. There are many advantages that it can offer to us and our clients. MT is a useful and extremely productive tool if used the correct way.
Large multi-national companies such as Microsoft, Symantec and Caterpillar have used MT successfully for a few years now. These organisations are using MT solutions for making large volumes of text available to their global customers in their local language without involving any human translators in the process. The Microsoft Knowledge Base, which contains more than 200,000 documents in English, is a well-known example of a text repository where the number of machine-translated documents by far exceeds the number of those translated by humans. Cost alone would prohibit these larger companies using a human translator to translate the after-sales documents that would be useful to their international customers.
But it is not only larger companies that can utilise the productivity gains afforded by MT. A CAT tool already integrates MT and it can translate text very fast. Is it perfect? No, but the answer is to have a human translator proof the text and spot the mistakes. A human translator working without a CAT tool can translate approx. 2,000 words per day. The same translator can produce an additional 2,000 words if they work with CAT or edit a machine translated text. Human translators can produce a first draft of the material using MT and then refine and edit it by hand. The cost of translation is now looking more affordable for customers in the global marketplace.