Language quality assurance
Language quality assurance (LQA) is the structured human review of a translation against defined criteria before it is delivered. A reviewer checks accuracy, terminology, fluency and compliance, often recording errors by category and severity. It is the human quality gate that automatic scores feed into, not replace.
How it works
An independent reviewer, not the original translator, checks the work against a rubric, frequently an MQM-style typology, logging each issue by type and severity. The result is both a decision, pass or fix, and a diagnostic record of what went wrong.
Automatic metrics and confidence scores tell the reviewer where to look; LQA is where a qualified human makes the final judgement.
How SourceTarget uses it
LQA is SourceTarget's final human gate. Metrics and confidence scores prioritise what the reviewer checks, but a qualified person signs off against the agreed criteria before anything is delivered or published.
Language quality assurance compared with Proofreading
| Language quality assurance | Proofreading | |
|---|---|---|
| Checks against | The source and defined criteria | The target text alone |
| Covers | Accuracy, terminology, fluency, compliance | Spelling, grammar, surface errors |
| Done by | An independent qualified reviewer | A proofreader |