Cleaning up the translation memory
With clean translation memories, you can get more out of every translation project.
With oneCleanup, oneword cleans up your TM database – for less rework, more consistent results and a solid foundation for artificial intelligence.
Our services at a glance.
What a well-organised translation memory can do
Translation memories grow with every project. Over the years, alongside useful translation outputs, sources of errors also accumulate: Segments where the same source segment is translated differently; phrases that are long out of date but are still suggested as matches; or sentence fragments where the source and target texts do not match.
Cleaning up the translation memory makes a noticeable difference. Translators work with clear TM matches that they can rely on. Fewer corrections are required during and after translation, and AI systems and machine translation are provided with a database on which to deliver reliable results.
A clean translation memory is not a one-off maintenance task, but an investment in every future translation and review process.
What we cover
Review and analysis
We analyse your TM to identify duplicates, fragmented content, outdated phrasing and structural weaknesses. You will receive a detailed analysis of where there is potential for optimisation and which steps will have the greatest impact.
Clean-up and consolidation
Duplicates are removed, TM matches are cleaned up, and outdated segments are specifically revised. We work with your existing file format, regardless of the manufacturer. With our oneCleanup service, we combine automated checks with human expertise and partially automate data-intensive tasks.
Structuring and handover
You will receive your cleaned-up TM database in the file format required by your systems and processes. This can be integrated directly into ongoing translation projects. For processes powered by artificial intelligence, the processed data set forms the basis on which machine translation systems and large language models deliver reliable results.
Ongoing maintenance
Clean-up is not a one-off process. We continue to support you even after the project has finished – with ongoing maintenance of your translations and clearly defined processes, ensuring that your translation memory grows in a meaningful way and remains reliable in the long term.
Here’s how we do it:
1
Review
Together, we’ll get an overview of your tools, your translation memory repository and what you hope to achieve through the clean-up process.
2
Analysis
Your collection is systematically checked for duplicates, fragments and form discrepancies. You will receive a results file and our recommendation for clear prioritisation.
3
Set priorities
Together, we’ll decide which areas to tackle first – without disrupting day-to-day operations.
4
Clean-up
We review the existing translation memory in a targeted and semi-automated manner. Initial results can be utilised at an early stage to have an immediate positive impact.
5
Handover
You will receive the cleaned-up data in the file format required by your systems, ready for use in production.
Our services at a glance.
When is a TM clean-up useful?
It makes sense to carry out a clean-up if the amount of revision required after translation output remains consistently high, even though a translation memory is in place. Also if AI-supported processes or machine translation are to be introduced, as these require a reliable database. If the TM database has been populated over the years by different individuals and service providers, there is usually considerable potential for cleaning it up. In all these situations, oneword takes care of the clean-up – in a structured and efficient manner, without disrupting ongoing projects.
FAQs
What exactly is revised during a TM clean-up?
Duplicates, fragments of translations and outdated phrasing are systematically revised, and the database is structured in such a way that the reliability of search results is restored. Exactly what we’ll be tackling depends on your existing corpus and your objectives – we’ll discuss this during our initial discussion.
Can we carry on working whilst the clean-up is in progress?
Yes. We plan work packages in such a way that ongoing projects are not disrupted. The first clean-up sub-sets are usually available at an early stage and can be put to use straight away.
Which computer-aided translation tools and file formats are supported?
We operate independently of any particular provider and support standard computer-aided translation tools and export formats. We will discuss what information you can provide us with during our initial meeting.
What is the difference between TM clean-up and terminology clean-up?
Both data sources have different structures and requirements. Translation memories are optimised for segment and match quality. Terminology databases are consolidated in such a way that terms, usage rules and entry logic are clear and unambiguous. Both can be commissioned separately or in combination. For more information, see the page on Terminology clean-up.
How can you tell that the effort was worth it?
Fewer queries, less rework, higher translation match rates and more consistent translation results. Following the clean-up, we provide specific key performance indicators that make it possible to measure the effect.
Further information about our technologies.
Let’s discuss your next project!
Please use our contact form or give us a call:
Make an appointment now:
