Quality Signals You Can Act On: The Latest from GlobalLink Translate & Review

Quality translated content is guided by context: what the glossary says, how the finished file will look, and what quality checks have already flagged.

This GlobalLink Translate & Review release adds to that context and brings more of it into one place. On top of the translation memory, terminology, and automated QA checks you already use, you can now filter machine-translated segments by their quality estimation (QE) score.

Here’s what’s new.

What Is GlobalLink Translate & Review?

GlobalLink Translate & Review is a browser-based translation review environment. It draws on data from GlobalLink TMS or runs as a standalone tool, working with TMX- or XLIFF-compliant translation memories and TBX-compliant terminology data.

What’s New in GlobalLink Translate & Review

Save time with quality estimation scores

Every machine-translated segment carries a QE score. If you post-edit at volume or run content review cycles on machine-translated material, the two controls below are worth setting before your next project.

Filter by MT QE score

Set a score range anywhere from 0 to 90 and the editor surfaces only the machine-translated segments that fall inside it.

Auto-lock high-confidence segments

Content segments scoring above your range can be locked automatically. With high-scoring translations settled, you can focus your time where it’s needed the most and deliver faster.

Read your machine translation quality checks in one report

Results from Legacy Transcheck and NextGen Transcheck now arrive together in a single, clean HTML document: the Combined Transcheck Report. With the power of NextGen’s fully customizable engine, false positives can be significantly reduced, so more of your review time goes to segments that need editing. Additionally, if the AI QA Agent is enabled, the agent’s results will also be included, making an even more comprehensive quality intelligence report.

Take Bilingual Review offline into Word and bring it back

Translators and reviewers asked for offline editing, and Bilingual Review delivers it. Work in Microsoft Word when it suits you, then bring your changes back into the editor with locked segments intact. Chained files are also exported together as a single document, making the whole review process easier to manage.

Next Steps

If you’re an existing GlobalLink Translate & Review user:

• Set a QE score range that matches how your team triages machine translation and run a file through it.
• Open the Combined Transcheck Report during your next review.
• Export a file with Bilingual Review, edit it in Microsoft Word, and import your changes back.

New to GlobalLink Translate & Review? Visit the GlobalLink Translate & Review page to see how translation, review, and quality estimation fit into one browser-based environment, or get in touch about a pilot for your team.

Quick Answers

What is quality estimation?

Quality estimation (QE) is a method for predicting and scoring the quality of AI-translated content. In GlobalLink Translate & Review, you can filter segments by their QE score, and segments scoring above a threshold you set can be locked automatically.

What is a QE score?

An QE score is the confidence value quality estimation assigns to a single translated segment, with higher scores indicating stronger AI output.

What is machine translation post-editing?

Machine translation post-editing (MTPE) is the work of reviewing and correcting machine-translated content. GlobalLink Translate & Review supports MTPE with translation memory, glossaries, and QE scores.