New Approach to Terminology Management

Blog post from Oct 05, 2017

Start Small, Then Expand

Time and again, I meet customers who run into various problems in connection with their internal terminology projects:

  • There is no budget for a dedicated terminologist, "somebody is taking care of it".
  • The terminology team comes together every one or two weeks (or longer intervals) and sometimes it feels like it hardly makes any progress.
  • Terminology is being collected, but nobody knows how and where.
  • The more extensive the terminology database gets, the less even the reviewers know the preferred terms from memory.
  • Every day, texts full of deprecated terms are written outside the technical documentation department.

All of this can be quite frustrating.

A terminology check, i.e. the actual use of the available data, is introduced only after the terminology has been fully entered. Instead, the processes are arranged as a waterfall model: Everybody is asked, and an overall structure must then accommodate the various requirements. Iterative approaches seem to be rather uncommon. Perhaps a repository may already exist, but might first need to be purged of wrong terms.

As a result, terminology is collected for years, but remains virtually unused. In terms of lean production, the terminology created in this way would even qualify as "waste". When actually will the terminology be complete or adequately collected? Many companies are waiting for the day that will never come ...

Terminology is alive and should be exposed to the reality check, i.e. it should be used in as many departments as possible.

Would it be possible to address these issues? What could an alternative approach look like?

Talking about lean production: You should start making efficient use of your terminology as soon as possible. You should use whatever you already have, and let the terminology live.

Terminology Check Even with an Incomplete Database

If you know numerous deprecated terms that authors use again and again, you should immediately start preventing them from being used. In this way, you will promptly achieve a quality improvement. Suddenly, the terminology database will no longer be mere "waste".

Even if the terminology circle should later adopt different preferred terms, deprecated terms will have been avoided in the meantime. The next terminology check will identify the status change in real time.

Don't get me wrong. I'm not trying to discourage terminology circles. To the contrary – coordination rounds are necessary and important. However, use of terminology that has already been approved should not be impaired due to the work that is still in progress.

Don't Be Discouraged by Missing Translations and Attributes

Terms in foreign languages are irrelevant to a terminology check in German. Don't be discouraged by missing translations and attributes that are irrelevant to the check.

If Necessary, Just Use Excel

If the budget is too tight, simply continue using Excel. All terminology checking programs I know of are capable of importing CSV files.

In time, the users will doubtless ask for a terminology database, e.g. due to the need for research options, ways to make proposals, or co-determination options.

Ensure Use of the Terminology

The more texts are checked for correct terminology, the higher the added value created by the terminology database. Therefore, onboard as many authors as possible. Encourage them to use the terminology database. If possible, get the backing of the management.

Efficient Terminology Check

Of course you could leave the terminology check to the authors and reviewers. For two important reasons, however, that wouldn't be advisable. Firstly, even the best reviewers do not know all preferred and deprecated terms of an extensive terminology database by heart. Secondly, a manual check is very costly.

An automatic terminology check within the scope of an author assistance tool eliminates the need for accessing the terminology database or an Excel file directly.  The author can work in his accustomed editor and does not need to exit his software in order to do manual research. This minimizes the overhead for the author. Tests conducted by Congree have demonstrated that without an author assistance tool, the terminology research and correction takes 13½ minutes, on average, for 100 sentences. With Congree, the overhead can be reduced to 2½ minutes. The calculation of the time input is based on the following parameters:

  • Average number of potential term candidates per 100 sentences: 30.26
  • Average research overhead per term: 20 seconds
  • Average correction overhead in the case of errors: 15 seconds
  • Average notification occurrence: 30%

Conclusion

Don't allow yourself to be frustrated by your current terminology situation. Start using your terminology in your enterprise – it's worth it. 

tl;dr

It is important to decide collaboratively about the collection of corporate terminology. However, use of terminology that has already been approved should not be impaired due to the work that is still in progress. Even a terminology check on the basis of an incomplete terminology database leads to text quality improvements.