How Machine Translation and Post-Editing Can Help you Reach the Market Faster
by Leandro Nunzio
Today’s fast-paced world is strengthened by technology and the Internet, producing more and more information at an increasingly faster pace. Therefore, it is important to have solutions that let you embrace these changes and, at the same time, make progress to be a step ahead of the competition.
At Go Global Language Consulting, within our translation solutions, we offer a tool that will allow you to have that extra content usefulness and speed up the information supply -the current common currency. This service is Post-Editing of Machine Translation (MT).
Machine Translation, or MT, is a utility that has been gaining strength during many years in the fields of translation and Computer-Assisted Translation tools (CAT tools), but it has been resisted by companies in general, and mainly by translators themselves, due to a number of well-grounded reasons.
The majority argues that working with MT is more time-consuming than a direct translation, because even though it saves time at the moment of generating the translated content, editing time of such generated content is greater than with a direct translation.
This could be true at the beginning, when the machine translation processing engines and servers were simply statistical, and the CAT tools technology was poor.
Today, many companies like SDL, memoQ, eBay, Google and Amazon offer enhancements in the tools processes and more independent programming, which allows programs to adapt the content they spread based on the context, to learn from the information being included into the translation databases, and to facilitate the expansion of post-editing use to other content-applied fields, besides translation itself.
For example, recent developments in MT enable customization of the use of the content according to projects, customers, fields (i.e. legal, medical, technical, etc.), and MT can be programmed to recognize when certain error is frequent and when a context is similar or equal to another, so that it can learn from that and recognize the range of instances, adapting content according to given settings.
Then, while its usage is enhanced, the tool works more and more predictably, and in turn, it learns more while it adds more content to its operation interface. Consequently, it offers more precise and better results that affect the working statistics of the finished documents and their preparation.
Viewed from another perspective, if the tool works better and its automated translations are more accurate, this implies the content analysis will be more precise, providing a more efficient quality estimate (so you will know with greater accuracy how much it will cost you and when it might be finished), and this will also allow the company working with you to deliver the end product to start using its own resources more efficiently in the efforts of the job and content production.
We are not saying the translator will be replaced by MT and, therefore, you may be able to buy a software capable of translating all the produced content in minutes and in a perfect way. That’s not the case. In fact, we will always need translators to edit the content, making sure it sounds natural and accurate, in terms of context and concept. The emotion, intention and myriad of meanings of a sentence or text always need to be properly interpreted and organized by a human being. However, post-editing of content with machine translation will feature a hybrid aspect to produce content in the target language, which brings the best of both worlds: advanced man-made technology, and the human being who operates it and guides it through the correct path -all for its own benefit.
To know more about Go Global and how our services can help you, contact us and a Client Solution specialist will be able to assist you.