Translators working with CAT tools are definitely familiar with term bases. But various genres of translation project imply different appropriation of term base use. In projects that emphasize fluency and style, like literary translation, film subtitling and so on, the application of term base may somewhat be not that urgent. However, term bases will so much come in handy for projects in which term consistency is highly demanded, like the translations of manuals, scientific journals, annual reports, and the likes.
However useful it is, you might often find cases where your term base doesn’t work as expected. For instance, it fails to identify words or phrases with the equivalences set. It’s bothersome, of course. But what’s more is it’ll affect the consistency of the translation and can be a serious problem for the translation where word-to-word references is crucial, like in manuals.
This undesired outcome usually happens when we input the words or phrases with inconsistent formats, like different capitalization, different separator between words in phrases, and so on. This commonly occurs when the entries we set in the term base incorporate long words or phrases. The longer the entry, the more likely this happens. Thus, it is better to input shorter words or phrases. In case of phrases, if possible, you can try to input the composing words instead of the whole long phrase.
Another cause of the term base’s failure to identify matches is typo. To anticipate this, we can select ‘fuzzy’ in the ‘Matching’ setting of the term base. Applying this setting, the term base will still detect the mistyped words, provided that the similarity with the entries is no less than 80%.
To make your term base effective, it is important to know exactly which words or phrases to put in as entries to then set the definite equivalents for them. First of all, you need to read your source documents to identify the frequently occurring words or phrases. In scientific journals, for instance, you can be sure to input the keywords or the other definite variables to the term base for they will surely appear many times in the whole text. Conversely, it is not necessary to input words or phrases that allow flexible renditions. You’ll find the presence of warning marks, because the tool thinks that you’ve made a mistake by not following the term base, unnecessary and only interfere with your working pace. If you need to record any of the flexibly translatable words or phrases, we can make use of the translation memory instead.