Most companies find that developing glossaries or terminology bases for machine translation is not as easy as it would appear. Here's a few things to think about and account for when developing a Glossary.
1. Know who are you developing the glossary for.
More and more companies are turning to machine translation systems to solve their translation challenges. Today's machine translation systems increasingly use AI to produce custom translations. Some machine translation systems will now allow users to produce custom translations with the aid of an interactive glossary. With an interactive glossary, machine translation terms are automatically replaced with terms from a user's glossary. Improving translation quality. However, preparing a glossary for a machine translation system is very different than for a human translator.
2. Only include non-ambiguous terms that have only one meaning.
Only include terms that are used in one way. Some words have two meanings. For instance, the term "clutch" can refer to a woman's purse or a clutch in a vehicle. In French the translation for assembly can be a verb (to assemble) or a noun (an assembly of parts) depending on how it's used.
3. For machine translation systems provide terms in multiple word shapes.
For machine translation systems you'll likely need to provide terms in multiple word shapes ( ex. - eat, eats, eaten), including: lower case, initial cap, all caps, singular and plural shapes as well. Machine translation systems require a 100% match with the term in the glossary and the files that get translated using the glossary.
4. Know the difference between a glossary and a translation memory.
A Glossary typically contains key words like; brand names, product names, acronyms and compound words that create one term. They are used to help produce consistent translations.
A Translation Memory is a bilingual (source and target language) repository in TMX file format. It consists of text that has been translated from one language into another, and it is stored on a server for future reuse. A translation memory consists of sentences, phrases and segments, whereas a glossary consists of key words, names and acronyms.
Glossaries and translation memories can deliver nearly instant quality improvements and reduce translation costs and effort.