In the localization industry, most translation memory systems (TMS) available for professional use are equipped with a host of automated quality assurance tools and functions. From basic Word-like spell checks and tag verification, to translation completeness and comparative analyses of names, numbers, and formats, translation memory software is designed to automate as many quality assurance tasks as possible.
To further complement the localization QA process, it’s important to integrate a multilingual glossary into a TMS when performing post-translation quality assurance. This is because, without a functional termbase, translation memory systems have no way of verifying the use of approved terminology between languages and across documents. As mentioned, a TMS enables translators to reuse previously translated content. The ability to leverage content, however, does not account for fuzzy matches in which only a portion of legacy content can be reused, nor does it account for completely new text that must be translated manually. Because of this, there is no way to automate the process of terminology QA without the use of a multilingual glossary.
If translators integrate a client’s termbase into their translation memory environments, however, then they can check terminological consistency automatically. Using a tool called a Term Checker, translators can run a multilingual glossary against the translation memory, thereby ensuring that only approved and preferred terms were employed in a given translation. In this same process, the TMS can also flag terms that deviate from the approved list of terms in the termbase, thus completing the terminology quality assurance process with a one-two punch: both preventing the use of deprecated terms and ensuring the use of authorized terms. Because many multinational corporations have a broad scope of products, developing product-specific (or even version-specific) glossaries further guarantees that the right words are used in the right place.
In practice, this function is useful because it guarantees complete linguistic consistency between documents, which is what terminology management is all about. When consumers sift through a product’s Online Help, for example, they rarely do so sequentially. On the contrary, owing to the increasingly powerful, Google-like search functions employed by most websites, users can jump directly to the section of a manual that most applies to the problem they’ve encountered with a product.
Because of this, it’s not uncommon for a troubled consumer to simultaneously reference, say, page 12 and page 253 of a single manual within the course of a few minutes. This is not a concern with traditional paper manuals because, if consumers were to read a user manual cover-to-cover, it’s highly unlikely that they would notice that, for example, a staple-removing device is referred to as a staple remover on page 12, and then as a destapler on page 253, because of the time spent in between. But this is a concern with online Help menus. Because of the ability to jump through online documentation, if a consumer encounters such a discrepancy within the first five minutes of investigating how to effectively remove a staple from his forehead, his subsequent frustration with the manufacturer would certainly be understandable.
Likewise, translators cannot be expected to notice a similar disagreement in terms without the ability to automatically jump through the content of a translated document and measure translation consistency against the source. Given the right tools—and a carefully prepared glossary—they can do so efficiently and with minimal room for error. The ability to hunt down discrepancies drastically improves the overall quality of a translated document and, by association, inspires confidence in the manufacturer.