The idea of using computers for translation is not a touch&go, it has been touch&stay since the 1950’s. It was first proposed by A.D. Booth in 1946 and in 1954, a completely automatic translation of more than 60 sentences from Russian to English took place in Georgetown as an experiment.
In 1985, Spain began a research funded by the three companies IBM, Siemens and Fujitsu. Since 1998, the University of Alicante develops the following systems of automatic translation for Romance languages: interNostrum (Spanish/Catalan), Traductor Universia (Spanish/Portuguese) and Apertium (Languages from Spain/other Romance languages).
Today, automatic translation is a need we face due to the large volume of information that the market processes. As so, human translation can largely extend the processing times of translation projects. Therefore to reduce these times, during an automatic translation the source is translated automatically and its results are revised by the post-editor, who in turn becomes the first human resource of the translation. The job of a Post-editor is to revise all mistakes that developed during the automatic translation and then hand down their work to the second human resource in the translation project, being the editor.
Fortunately, the automatic program can be adjusted by a linguist. For example, the names of companies can be blocked to avoid their translation. Another pre-editing configuration can be a glossary. This way, the system will be set up in a way in which it can read a segment and translate it consistently throughout the document, and according to the linguist’s preference. This improves and provides consistency to the automatic translation product. However, the automatic product still depends widely on human support to attain fluidity and grammar correction in the final text.
To read more about automatic translation tools, we recommend the following articles found in our blog: