QA process and QA tool
The above three tools, i.e. glossary, translation memory, and style guide, can serve well in helping maintain consistency, but to produce a high quality work widely accepted by the end users, it has to be ensured by a QA process from Translation, Editing, and Proofreading (TEP). We rely on a complete process instead of a particular individual to guarantee the quality.
At each stage of the process, a general checklist is provided and the assigned resource is supposed to strictly follow the checklist before handing off to the next stage. For each individual project, an additional project checklist is created according to the specific requirements of the client.
The project manager is supposed to strictly follow the project checklist before delivering the project. While a well defined process sets the keynote of the quality of the service and delivery, the personnel engaged in performing the QA role is the key to the success of the process and the project.
The key QA personnel have to be qualified, experienced and knowledgeable in its field and most of all, be deeply rooted in the sense of responsibility. That’s why we are saying it’s a forever challenging task – right people do the right things in the right time – which is also what we have been pursuing.
Apart from the eye checking by separate linguist, you can also take advantage of some QA tool such as Xbench to catch those errors that are difficult to spot by eyes. By defining your current translation files as ongoing translation in your Xbench project, you can run the following QA checks:
1. Find untranslated segments
2. Find segments with the same source text and different target text
3. Find segments with the same target text and different source text
4. Find segments whose target text matches the source text (potentially untranslated text)
5. Find tag mismatches
6. Find numerical mismatches
7. Find double blanks
8. Find terminology mismatches against a list of key terms
9. Execute user-defined checklists.
10. Checklists are user-defined searches that you can run in batch against your ongoing translation. For example, with checklists you can search for banned words or typical translator pitfalls.