The magic of batch changes in termbases

…and why MultiTerm from RWS Trados is our Go-To!

But one step back – what does it even mean to “modify” a termbase?

As computational linguists at berns language consulting, working with extensive terminology and termbases is an integral part of our daily business. From setting up termbases from scratch, converting formats to executing large-scale modifications and mergings – we can handle it all with linguistic and technical expertise. To provide these services, we rely on our Data Toolkit, a self-developed Python framework. Additionally, our strategic partnership with?RWS?Trados?enhances our capabilities, enabling us to deliver even more robust and efficient language solutions.

“Modify” means to clean up a termbase at the structural level. The structural level refers to the metadata – in comparison to the content level, which refers to the terms and concepts themselves.

Concrete examples for cleaning potentials on a structural level are renaming metadata fields, changing the level of fields, mapping field values or merging duplicate entries. Also, building a termbase out of non-termbase terminology data (e.g. Excel lists) or enriching an existing termbase with more data are special cases of modification.

Why modify a termbase at all?

Use case: Enrich a termbase with additional terminology data

Here is an example of a use case in which we added terminology equivalents in around 50 languages to an already highly complex monolingual database of a German car manufacturer.

And how do we do it?

In an object-oriented Python framework, files are read and their content, along with metadata, is represented within Python objects. All changes and manipulations are performed directly on these objects, ensuring a clear and organized workflow. Finally, the modified termbase is saved to a file in the desired format, maintaining the integrity of the processed data.

Excuse me – what objects?

Object orientation is a programming strategy that allows you to model real-world entities as objects in programming, making it easier to organize and manage complex systems. In our Data Toolkit the objects recreate the hierarchical structure of a termbase, i.e. the levels concept, language, term and various meta data fields.

As there are various formats of a termbase, we had to decide what our objects should look like. Since the standard on the market is?MultiTerm?XML, a format developed by RWS Trados, we have structured our objects accordingly. We have even gone so far as to convert all other formats into this format using specifically written parsers/readers. Theoretically, you could also write parsers/writers in this way after changes have been made to the objects, in order to be able to transfer them into various formats. However, as most of the widely used terminology management systems on the market are compatible with MultiTerm XML?as an import format, we have generally not needed to take this step before. Furthermore,?MultiTerm?is often utilized as a preferred system.

Partnership with RWS Trados

Trados is part of RWS, a world-leading provider of AI-powered language and content technology and services, working with 88 of the top 100 global brands. As a leading manufacturer, RWS Trados sets standards in localization. The tool solutions are widely used and many of our customers rely on RWS Trados solutions. Trados delivers end-to-end solutions for everyone involved in translation – whether a new translator looking to use their first computer-assisted translation (CAT) tool, a language service provider looking to collaborate more effectively with clients, or an organization looking to manage large volumes of multilingual content.

As independent partners, both blc and RWS Trados benefit from exchange, expertise and close cooperation!