Localising Prezi presentations using Computer-Assisted Translation (CAT) tools

Localising Prezi presentations using Computer-Assisted Translation (CAT) tools

For a lot of presenters who have always wanted to stand out (not to mention a whole generation of Star Wars fans dreaming of piloting Anakin’s Eta-2 Actis-class interceptor – or at least his Tatooine podracer – until everyone involved gets dizzy and faints), Prezi is the best thing that has ever happened to them: a funky, cloud-based, collaborative presentation authoring environment which can be accessed on a wide range of devices and comes with a ton of zooming in transitions. Perfect!

What is not perfect, however, is that, when wanting to reach global stardom, these presenters together with the linguists they employ in order to translate their presentations into a variety of languages have a problem: Prezi doesn’t export the translatable content for easy localisation and Quality Assurance. The only supported workflow is a manual one, involving overwriting/copy-pasting translations over original text.

Until now I haven’t been able to find a guide on the most effective way of using CAT tools to localise Prezi presentations, so here is the quickest Prezi localisation process I could find after a bit of playing with a range of CAT tools: a 30-minute video that takes you through all the steps both presentation authors and linguists need to take in order to achieve a consistent and good-quality product (I have embedded the video below, but if you prefer to watch it in YouTube, here is the direct link). The CAT tool that offered the most painless process in this scenario was Memsource, but you can, of course, also use other CAT tools to localise the .xml file which contains the presentation content.


I hope these quick pointers help and I’m looking forward to your comments on Twitter & LinkedIn.


1. In the video I should have made it more clear that, unless you need the title of the off-line Prezi presentation to be localised, too, you don’t need to deal with the .json file content (you may as well delete altogether from the Memsource project).

2. If the Prezi contains images with embedded text (really bad authoring practice, but unfortunately still fairly common), that text will not be extracted by Memsource – no CAT tool will extract it, in fact, but memoQ goes furthest at present in allowing you to create localisation packages which also involve images. You will therefore need to find the images in the Prezi ‘image’ or ‘swf’ subfolders and deal with them from there.