Translation jobs now have glossary and reserved word functionality built directly into the preprocessing of content significantly improving the speed and accuracy of our translation services
The great thing about having to process an increasingly diverse range of translation projects is the ability to try different ways of implementing some key translation features we have known we needed to implement. Glossaries and protected terms are fairly standard requests from any organisation that does consistent translation work. As a quick review below is how we define both options.
The Glossary: A glossary is essentially a grouping of word pairs between a source language (normally the operational language of the organisation and the languages they translate into). Below is an aexample of a glossary between English (source) and Brazilian Portuguese (Target). The glossary is a grouping of terms that the client wants to expose to the translator as a guide to how they want that term/phrase translated.
Reserved Words: are terms that an organisation does not want translated. Reserved words are normally words/phrases like brand names, product names, acronyms that have a specific meaning across language barriers.
From a translators perspective the briefing is simply, reserved words you dont change but view a glossary item as a strong guide from the client (but you have the ultimate decision on what to use).
The challenge for our team was how best to implement this very important functionality whilst ensuring we dont overly complicate the translator workbench or make it difficult (time consuming) to identify these terms in the content.
After trying multiple different approaches we actually found the best approach was to tag the glossary and reserved terms in the preprocessing phase and highlight the relevant terms inside the workbench (note best approach and most technically challenging went hand in hand here). If a glosssary term exists the translator can see them identified as <glo WORD glo> and a reserved word by <res WORD res>. It took some tweaking to get the preprocessing accurate were multiple words (i.e. phrases) exist but its near 100% accurate now.
The results so far have been beyond our expectations. By identifying the terms in the workbench in context (i.e. inside the text to be translated) the translators can actually quickly make a call on whether the term and phrasing is appropriate and the quality of the translations has improved significantly as they adhere more closely to the clients brief.
We just rolled out these new features to all clients who use our translation services – if you want more information send us an email to firstname.lastname@example.org