Cleaning up terminology

Accurate terminology as a reliable foundation for translation, editing and AI.

With oneCleanup, oneword systematically cleans up your terminology database in multiple languages to ensure your specified terminology is used consistently throughout every process.

Arrange a consultation now

Our services at a glance.

The benefits of a cleaned-up terminology database

A clean terminology database reduces the need for queries, cuts the time spent on corrections, increases the hit rate in the translation memory and provides AI systems with reliable guidelines. Analyses from our projects show that: Some databases contain only 20 to 30 percent of actively used terminology. Around half of all corrections made in both the source text and the translation relate to terminology. A streamlined repository reduces this workload in the long term.

By getting your terminology data in order, you provide everyone involved – in-house teams, external service providers and AI systems – with a common, binding foundation.

Arrange a consultation now

What we take care of

Analysis and prioritisation

You’ll be given a clear indication of which terms are actually in use and where it's worth investing your time and resources. To do this, we cross-reference your existing terminology database with the latest corporate texts and translation memories. Our oneCleanup service then combines automated checking steps with terminology expertise. This way, you’ll know exactly where action is needed before resources are allocated.

Structural reorganisation and database optimisation

Your database will be structured to act as a single source of truth for everyone who accesses it. We revise fields, consolidate scattered information and ensure that terms are accompanied by clear usage information.

Clean-up of entries and removal of duplicates

You’ll have a consolidated database that you can build on. We revise duplicates, orphaned entries and inconsistent or outdated terms, and merge existing glossaries and lists. The cleaned-up termbase can then be managed centrally, for example in oneTerm – ensuring consistent results in day-to-day use.

Multilingual clean-up and equivalence checking

We check equivalents in all the required foreign languages, add any missing terms and ensure usage information is clear and consistent across all languages. This ensures your specified terminology is binding in every language in which you communicate.

Arrange a consultation now

Case study:
Terminology clean-up at HOMMEL ETAMIC

How oneword analysed, structured and cleaned up over 28,000 terms in 20 languages for a global industrial company – and which steps had the greatest impact.

Download case study

Here’s how we do it

Initial consultation

We’ll get an overview of your database, the tools you use and your objectives.

Analysis

We determine the proportion of the database that is in active use and identify structural issues and potential for cleaning up.

Setting priorities

Together, we decide which areas to tackle first, without disrupting ongoing projects.

Clean-up

Entries are revised, the database structure is adapted and equivalents in the required languages are added.

Handover

You will receive the cleaned-up data in the file format required by your systems, translators and AI applications.

Quality control

You will receive the cleaned-up data in the file format required by your systems, translators and AI applications.

When terminology clean-up is the right choice

Terminology clean-up is the right place to start when terminology data has grown over the years and it is either unclear or poor in quality. It’s also a wise choice where specified terminology exists, but the amount of correction required after translation remains consistently high.

A cleaned-up database is also the ideal starting point for the introduction of machine translation or AI-supported processes. Terminology can, for example, be integrated into machine translation and AI systems in the form of a glossary. However, if a glossary contains superfluous entries or ambiguous rules, this significantly reduces the quality of the machine-generated output. Anyone wishing to incorporate terminology into glossaries for AI systems should therefore start by performing a clean-up. The same applies before switching to a new tool or system, because a clean database makes migration considerably easier.

Arrange a consultation now

FAQs

What’s the difference between terminology clean-up and terminology creation?

Terminology creation involves building a terminology database from scratch or systematically expanding it. Terminology clean-up builds on existing terminology and resolves inconsistencies, duplicates and structural weaknesses.

What’s the difference between terminology clean-up and TM clean-up?

Terminology databases are consolidated in such a way that terms, usage rules and entry logic are unambiguous. Translation memories are optimised for segment and match quality. Both can be commissioned separately or in combination. For more information, see translation memory clean-up.

How can I tell if my terminology database needs to be cleaned up?

Clear signs of this include frequent queries of terms and meanings, and a consistently high volume of corrections required after text creation and translation, despite the existence of a terminology database. Other indicators include: terms needing to be cross-referenced across multiple sources; synonyms having been created without usage information; or entries existing only in a foreign language with no reference to the source language.

What happens to entries that are no longer actively in use?

Inactive terms are initially flagged as such but are not deleted straight away. This means that inactive records remain accessible, whilst it is possible to filter by active and inactive entries. After a specified period, inactive entries can be selectively removed.

Once the clean-up is complete, can I ensure that the repository remains organised and up to date?

Yes. An optimised structure ensures that the database will be populated more efficiently in future. We also recommend converting the clean-up into a continuous process, in which the database is maintained and expanded as translations are carried out. Guidelines and defined workflows ensure that responsibilities are clear and new entries are created consistently.