Machine Translation has long been used as a tool for ‘gisting’, to get an approximate translation for self use. Professional translators are often not willing to accept projects which entail post-edition of Machine Translation. Typical excuses include: “the machine translation quality is too poor” or “it’ll take me less time to do it from scratch.”

But machine translation has come a long way in recent years, a long way from the rules based translation systems we’re used to. The professional community is gradually being convinced that MT has its place in the future, when used with efficient post-editing the MT process has several benefits:

  • the ability to deliver projects faster. In comparison, a professional translator using a computer aided translation memory technology can translate approximately 2,500 words per day. A very fast translator with CAT and used to the subject domain can translate more, but not a huge amount more. If for example a translation agency gets a project of 200,000 words then it would take about 80 translator days to deliver the project. So if you assign the project to 8 translators, the project would take 10 days to complete. But what if you need to deliver the 200,000 words in 3 days? Assign more translators? The process and translation consistency will become more difficult. Using a process of machine translation with post-edition makes these types of volumes and delays possible, post-edition throughputs for a good reviewer is approximately 10,000 words per day to deliver high quality text.
  • Cost savings – machine translation with post-edition is often less expensive than classic human translation using computer aided translation memories.
  • If the machine translation and target language texts are well managed and integrated into a translation memory process, both quality and leveraging improves dramatically with increasing translated volumes. Machine translation relies on terminology management and quality input for training and customization which in turn guarantees translated text consistency. So, investment in machine translation will pay dividends over time.
  •  Using online machine translation systems hosted by large search engines and IT companies does not guarantee that you source and translated texts will not be reused in the public domain. Using machine translation services hosted and managed internally by your translation agency guarantees confidentiality.

An explosion of content being published on the Internet in the form of websites, blogs and social networks has created a dramatic increase in the amounts of potential texts for translation. The very nature of these web-based texts requires a new approach to offer faster and cheaper translation alternatives. Machine translation has it’s role to play in the translation of these ever increasing sources for there are simply not enough professional translators to keep up with all of this information and able to translate content at the pace required by users.