This is a guest post from Memsource. In their own words, “By leading the industry in AI-powered translation technology, we make localization easier, faster, and more cost-effective.”

Memsource and Amazon Translate are strengthening their partnership. You can now use Amazon Translate with Memsource Translate, Memsource’s machine translation (MT) management feature.

Many Memsource users share a common problem—there are so many MT engines to choose from that it’s a difficult and tedious process to decide which one is the best for translating your source content. Additionally, changes are made to the engines frequently and new language pairs are constantly added. In a Memsource internal assessment, we found that in over 70% of translation projects, customers aren’t using the most effective MT engine.

Memsource Translate is a new MT management feature that automatically selects the best-performing engine for your content, which makes it easier and faster to produce high-quality translations. It also includes the latest version of Memsource’s AI-powered feature, Machine Translation Quality Estimation (MTQE). This feature provides real-time quality scores for MT output before any post-editing is done.

We are very pleased to announce that Amazon Translate has just been added as one of the supported engines.

To use Memsource Translate with Amazon Translate, all you need is a Memsource account. When you first sign up, you receive free Memsource Translate characters to test out on your content.

How selection works

Selection is based on the language pair of the translation. The following diagram illustrates the process.

memsource

When you upload your content and select your language pair, Memsource Translate auto-selects the highest-ranking engine. You can then review your MT quality scores. As a result, you increase your translator productivity and reduce costs.

Looking forward

In the coming months, the content type (domain) of the text will also play a role in MT selection. The domain is automatically identified by using AI-driven technology. The domains range from Software to Travel and Hospitality. Future features also include adding your own trained or adapted MT engines to Memsource Translate.

To get started with Amazon Translate in Memsource, see Pricing. Let us know your experience.


About the Author

Jitka Hankova is a Product Manager at Memsource, a software company providing cloud translation technology, based in Prague, Czech Republic. Jitka is a translation technology enthusiast and has worked in the industry for 8 years.

 

from AWS Machine Learning Blog: https://aws.amazon.com/blogs/machine-learning/amazon-translate-added-to-memsource-translate-memsources-machine-translation-management-feature/