A terminology database (Termbase) is a collection of predefined terms related to a particular brand. They greatly simplify life for translators and proofreaders. When translating a new source sentence, translators get automatic translation suggestions for single terms that appear in the sentence from the Termbase.
An example of what might be in included in a Termbase is whether a brand hyphenates the phrase “time-to-market” or not. Automated prompts like these help brands to stay consistent and speed up delivery times.
A Termbase offers an overview of your brand terminology.
Termbases typically provide extra information about terms, such as approval and usage status, notes, the creation date and last edit date. It is often a rule that Termbase suggestions should be used; even if a translation memory match contains a different term, a Termbase overrules a TM.
Machine translation has come a long way in recent years.
The 1990s saw a development of statistical techniques for machine translation. By 2016, there were significant advances in the science behind MT, bringing rise to the era of neural machine translation (NMT).
Neural machine translation (NMT) is the current MT technology of choice. What makes NMT so popular is that the engine learns over time, and as it learns, it becomes more accurate. NMT produces the most accurate translations compared to any other MT technology available today.
The network used for machine translation consists of more than a hundred million virtual neurons – all inspired by the human brain. The more the engine is trained, the better the translation it produces. The engine can also be trained with customer-specific data from a terminology database or translation memory.
Free white paper: Optimise Content Workflows with Machine Translation.
Are you struggling to create more content, in more languages, in less time, and at a lower cost? Read this handy guide on how to streamline your content workflows with Neural Machine Translation.
MT is a rapidly changing area of research and development. LanguageWire’s data scientists and engineers use NMT technology based on open-source solutions, continually evaluating the rapid framework development and cross-evaluating output quality.
As the translation industry evolves and more companies adopt new technologies, custom NMT solutions are increasingly being used to optimise content workflows.