WebProNews

Google has announced the launch of the Google Translator toolkit, an editor designed to give translators an easy means of bringing the "human touch" to machine translation, which everybody knows is often flawed.

Michael Galvez and Sanjay Bhansali of the Google Translator Toolkit team explain how the toolkit works:

For example, if an Arabic-speaking reader wants to translate a Wikipedia™ article into Arabic, she loads the article into Translator Toolkit, corrects the automatic translation, and clicks publish. By using Translator Toolkit's bag of tools — translation search, bilingual dictionaries, and ratings, she translates and publishes the article faster and better into Arabic. The Translator Toolkit is integrated with Wikipedia, making it easy to publish translated articles. Best of all, our automatic translation system "learns" from her corrections, creating a virtuous cycle that can help translate content into 47 languages, or over 98% of the world's Internet population.

The Translator Toolkit is comprised of tabs for:

- Translation Search Results
- Computer translation
- Glossary
- Dictionary.

In addition to Wikipedia, the Google Translator Toolkit is also integrated with Google's Knol, and supports common document types like Word and HTML. Terminology and translation memory management are provided for translation professionals.

As users translate segments, Google saves the translations either to its shared translation memory, or to a translation memory of the user's choosing. The purpose of this is to save time in the future.

To move to different segments in a translation, users can click the source or translated segment on the editor, click the next and previous links on the edit box, or use Ctrl+J to move to the next segment and Ctrl+K to move to the previous segment.