If you want to save time and cost in translation, and minimise errors, then incorporating a terminology database is a must. In my last post, I looked at the benefits of using a database to manage terminology.
But they can be expensive and difficult to create, so our challenge was automating as much of the process as possible.
So how did we do this? We broke down the task of building the database into discrete steps. Then we looked at each one to see how this could be automated without losing quality, and where our linguists could easily add value to the data.
For the first stage, our world-class machine translation team came up with a solution. They fine-tuned the process of extracting potential client-specific terms (“term candidates”) from the client’s bilingual data, and the appropriate translation for each one.
Our success rate in extracting source terms was initially between 74 per cent (for less common languages) and 95 per cent for the major language pairs.
In other words, we can extract relevant and complete glossary entries from translation memories with a very high success rate.
These results were impressive enough to be published in a leading machine translation journal.
At this point we added a human touch to the mix, by building a simple interface to evaluate the glossaries.
The degree of evaluation can vary from a simple “yes” or “no” to a qualified terminology filling in comprehensive information on each source terms.
We can help clients make this decision as part of our consultation process.
Once the evaluation is complete, then we create the terminology database. It’s compatible with all high-end computer-assisted translation (CAT) tools, including Coach, our proprietary software.
So how does it work? When a translator is working using Coach, any term in the database will be highlighted with its suggested translation. If a specific term is found in the source content, but its translation isn’t in the target, then this will alert the translator with a “quality warning”.
We can offer some additional services to improve the database even further. This includes appointing designated terminologists for each language combination, who manage the terminology and ensure it’s always up to date. They can extend and improve the database, while clients can access a fully featured terminology management tool.
From building a basic glossary to creating fully-fledged advanced terminology databases, our software streamlines the process. The benefits are easy to see: more consistent, faster translations, and an improvement in quality.
And that’s just the start. As translations progress, both clients and linguists can contribute to the terminology database, adding and updating terms to continually improve the results.
*Picture credits: Vladru / Shutterstock.com