Will Google Finish Off Human Translators-govos

By , September 16, 2018 3:28 am

UnCategorized It is a well known fact that Google has invested heavily to breach the language divide through technology and computer aided translations. The company considers their computer translation tool as one of the key components of their global business strategy. The reason is quite simple: machine translation will open large markets for Google and it is a key element for their content monitization strategy. The Company now has expanded its language translation offerings to 41 languages including Turkish, Estonian, Galician and other languages. Beyond the business goals of Google, this new technology will certainly impact the way the world communicates. It is one step on the road to eliminate language as a barrier to communication. At the same time, the advent of machine translation technology has many translation service buyers asking: Will this technology render human translators irrelevant? The short answer to this question is no. At least for the foreseeable future machine translations are grossly inadequate to translate sensitive or complex texts. The reason is that the very nature of language is highly idiosyncratic and context driven. Computers are not well equipped to decipher meaning. Google’s approach is to compare huge databases of known translations to construct an approximate translation. This works reasonably well for some applications, but it is unreliable at best. As a language translation provider, we have tested translations done by Google’s translation tool in a few language combinations. Richard Ramos, President of LanguageTran, said "Google has achieved a remarkable level of accuracy in some of the language combinations. It is a great way of translating content to have a general idea of the original text. However, the intrinsic nuances of language are still trumping the ultimate goal of near perfect translations. I can see the day when we use tools like Google to improve productivity, but we don’t see that happening anytime soon. In some of the language combinations that we tested some sentences were perfectly translated. However, errors in terminology, grammatical constructions and subtle interpretations in large parts of the text make the tool unusable for applications which require accurate and well written translations." He added, "The time needed by a human editor to proof-read, edit and correct errors is still greater than the time required for direct human translation". Can we program a computer to write a new Shakespearean play? Don’t bet on it. We can use computers to write data driven text using mechanistic constructions, but writing is an art and translation is no different. Some translators are using machine translations to increase their work productivity. In our experience, it takes longer to edit a poor machine translation than to translate it from scratch. Computers have a long way to go before they will render human translators useless, if ever. About the Author: 相关的主题文章:

Comments are closed