Pre-translate a document

In this guide: Starting the translation process for a new document can be overwhelming as you realize there is so much work to do. Learn how to pre-translate your documents on Redokun using previously approved translations or machine translation in order to reduce your team's workload.

Pre-translation is a simple process that consists of 2 sequential steps:

Translate with Translation Memory: This step uses your previously approved translations to translate your document.
Translate with Machine Translation: This final step translates the remaining segments using translations generated by our MT vendors.

How do you run a pre-translate?

The pre-translation process can be run at any time.

Go to the Document Detail page and click on the Pre-translate button as shown here:

Choose whether you want to use Translation Memories only or Translation Memories + Machine Translations for the target language you want to translate.
Configure the options for Translation Memory and Machine Translation as needed. Note that the Machine Translation options are available only if "Translate with Machine Translation" is selected. Marking translations as 'Draft' allows you to invite your team members to quickly review and check translations that might need human attention.
Click on Confirm to pre-translate the document.

Translate with Translation Memory: options available

By default, Redokun prioritizes translations with the same surrounding segments, as these are generally more accurate. However, you have additional options for using Translation Memory with different context in previous documents by changing the option "Also use Translation Memory with different context in previous documents" to:

Yes, and mark all as 'Confirmed' (default)
Yes, and mark all as 'Draft' for further review
No, use only translations with the same surrounding segments

Translation Memories are typically more accurate than Machine Translation, so we recommend prioritizing them over MT-generated translations."

What is Context?

The term " context" simply refers to the text surrounding a particular text segment.

What is a 100% Match?

100% matches occur when a source segment you're currently translating is identical to another source segment your team has translated before.

In these cases, you can usually reuse the same translation with a good amount of confidence.

What is a 100% Match with Same Context?

This is same as the above but it takes the matching algorithm a step further.

A 100% Match with Same Context occurs when both of the following conditions are met:

A source segment you're currently translating is identical to another source segment your team has translated before.
The source segment in both the new document and the old document share the same surrounding segments (context).

This matching algorithm is especially useful if you regularly translate documents with repetitive sections. Brochures and product catalogs are a good example as they are likely to have a few parts that do not change as your team cycles through different editions. For instance, the company profile or the descriptions for a product line. With a deeper level matching algorithm, you can safely reuse your translations for the parts that are repeated across different documents. And they likely won’t require any editing at all.

Compared to a 100% Match, translations marked as 100% Match with Same Context have an even higher likelihood of being accurate and on-topic.

The following visual highlights which segments are considered in determining whether there is a 100% Match or 100% Match with Same Context.