| Subcribe via RSS 

Machine Translation in Practice

November 27th, 2012 | by LTD

Session LT-10 at ATA 2012
Presenter: Mike Dillinger
Reviewed by: Rubén de la Fuente

Mike Dillinger gave a very thorough introduction to MT, ideal for people who want to get more acquainted with this technology.

First interesting point is that MT and TM are not that different: they both re-use their stored linguistic resources to produce translations, but while TM stops at a sentence level, MT will go one step further and work at phrase level, thereby increasing leverage. Also, like TM, MT needs to be customized upfront with relevant translations in order to perform well.

MT opens a few career paths for translators: development of MT systems (engineers will need our linguistic expertise to be successful), post editing and turbo translating (© Mike). Let’s focus on the last 2, since they are closer to us now.

Funnily, PE can be more profitable than translating. The productivity increase achieved by using MT will enable you to translate more words in less time, boosting your hourly revenue and freeing up time to take more jobs (or to relax). This will only be possible with MT output of good quality, that can be re-used at least in 60% (i.e. you should only have to fix 40% of the job overall). It’s very important then to develop an eye to spot good MT output from bad MT output. A good trick is to check key terminology: if it is not translated right, most likely there’s been no upfront customization and the PE job will be a nightmare.

In turbo-translating, translators have an MT system of their own (be it commercial like Systran or open-source like DoMY CE) and take care of post editing too. This is the best scenario for translators: 1) they can tune the system to take care of any recurring errors they find in PE; 2) they get all the profit from using MT (whereas they only get a share of it when PEing 3rd party MT).

MT is a technology with plenty of potential, but if you want to use it your attitude needs to be glass half-full: appreciate the pros and don’t be put off by the cons.

Rubén R. de la Fuente has a BA in translation and interpreting from the University of Granada. He has over 10 years of experience in localization in various capacities, including as a freelance and in-house translator, reviewer, project manager, and machine translation specialist. He is currently taking a graduate course on computational linguistics. He has taught several courses and workshops about translation tools for the Universidad Alfonso X and organizations such as the Institute of Localisation Professionals, ProZ, and ecpdwebinars.co.uk. He has written articles on translation tools for ATA’s Language Technology Division. You can reach Rubén at rubo@wordbonds.es.

Leave a Reply