L07 Machine Translation

title: "Machine Translation for Technical Documentation" weight: 7 description: "Learn machine translation for documentation: tools like DeepL and Google Translate, post-editing workflows, quality evaluation, and when machine translation is appropriate for technical content." date: 2026-06-28 lastmod: 2026-06-28 tags: [technical-writing, localization] }

Machine translation uses AI to translate technical documentation, offering speed and cost benefits but requiring human post-editing for accuracy with specialized terminology.

In this lesson, you will learn machine translation tools, post-editing workflows, quality evaluation methods, and when machine translation is appropriate.

What You'll Learn

You will learn about MT tools like DeepL and Google Translate, post-editing best practices, quality evaluation, and appropriate use cases for machine translation.

Why It Matters

Machine translation with human review can reduce costs by 50 to 70 percent while maintaining quality for many documentation types.

def machine_translation_workflow():
    """Define MT with human review workflow."""
    return [
        "Source text prepared",
        "Machine translation applied",
        "Post-edit by human translator",
        "Terminology consistency check",
        "QA review",
        "Integration",
    ]

Teacher Mindset

Think of machine translation as a first draft by a very fast, occasionally confused assistant. It gives you 80 percent of the way there. The human editor provides the final 20 percent of quality.

Common Mistakes

1. Publishing Raw Machine Translation

Raw MT contains errors. Always post-edit.

2. No Terminology Management

MT handles technical terms inconsistently. Use glossaries to guide translation.

3. Ignoring Context

MT translates sentences, not paragraphs. Without context, translations can be wrong.

Practice Questions

1. What is post-editing? Human review and correction of machine-translated text to ensure accuracy and fluency.

2. When is machine translation appropriate for documentation? For internal docs, drafts, and low-visibility content. For customer-facing docs, use MT with thorough post-editing.

3. Challenge: Run a documentation page through DeepL. Post-edit the output. Compare quality with human translation.

FAQ

Mini Project

Take a documentation page. Run it through machine translation. Post-edit the output. Evaluate quality against a human translation of the same page. Document the cost and time difference.

What's Next

Human Translation in the next lesson.