Postredigering

Postredigering, or post-editing, is the process of revising a text that has been translated by a machine or partially translated, with the aim of restoring accuracy, fluency, and suitability for the target audience. It is commonly used in localization, content production, and multilingual communications to speed up translation workflows while maintaining quality.

There are two main levels of postredigering. Light post-edited translations (LPE) are revised to be understandable and generally accurate, with minimal changes to style or nuance. Full post-edited translations (FPE) strive for human-level quality, including natural language flow, precise terminology, and faithful representation of the source, suitable for publication.

A typical workflow includes assessing the source text, applying machine translation if used, editing for meaning, terminology consistency, tone, and readability, followed by proofreading and quality checks. Terminology management, style guides, and translation memories are often used to ensure consistency across content and projects.

Tools and standards commonly support postredigering. Computer-assisted translation (CAT) tools, MT engines, glossaries, and translation memories facilitate efficiency. Standards such as ISO 18587 provide guidance for post-editing of MT, while quality models like MQM or LISA help evaluate the revised text. Metrics such as post-editing effort, edit distance, and time per unit are used to measure productivity and quality.

Challenges include balancing speed with quality, maintaining consistent terminology, addressing domain-specific content, and ensuring data privacy. The outcome of postredigering is a target text that meets the intended purpose, along with a change record that can inform future MT improvements and workflow adjustments.