Using Glossaries and Other Techniques to Improve Computer-Assisted Translation

Human translation of large document collections is expensive and out of range for most cases. Simple machine translation can be comically inaccurate.

Anyone who has used Google Translate knows that computer software can be used to translate a wide range of languages, doing a decent job with many of them. At other times, the software does a poor, often comical job, providing all but incomprehensible text as a result. This is particularly true for many Asian languages including Chinese, Japanese and Korean.

Catalyst's Enhanced Machine Translation (EMT) service provides a cost-effective answer to this dilemma. Our language specialists work with you to prepare a glossary of key terms in a case. By integrating the glossary with our "best-in-class" MT translation system, we produce far more accurate translations. Save significantly by using EMT for your first pass, then pay for human translation of only the most critical documents.

Using a Glossary for EMT

Enhanced Machine Translation allows you to build a glossary that can improve translation results dramatically. When the computer software attempts to translate a word of phrase, it first checks the glossary to see if there is a preferred translation. That simple step can substantially improve translation quality.

We build glossaries in a number of ways. To start, Catalyst consultants extract key terms and phrases from the document population and submit them for hand translation.

A second step is to identify unknown words not found in the standard EMT dictionary and list them by frequency. We have a human expert translate the significant ones and add them to the case glossary.

As a third approach, we add English documents to the glossary because it can be used for comparative purposes by the system. The result is a more natural phrasing for the translation and improved comprehension.

The result is a secure case glossary that can be used during the review phase. While these translations are typically using for gisting--to get the essential thrust of the document--they can be helpful in determining whether a document is important or whether it can be dropped from the review.

On the Fly or In Bulk: We Provide Both Options

We support translation on the fly or translation in bulk for 38 languages into English. Your reviewers can click on a foreign-language document and see it translated within seconds. You can also designate multiple documents for translation and direct the process with a click. Translated documents are saved in the system for future access.

Using Human Translators to Supplement EMT

The perfect role for human translation comes when you have identified those few documents that matter for your case. At that point, we can provide certified human translators to perfect the EMT and to certify complete translations for court or a regulatory hearing. Cleaning up EMT is less expensive than starting from scratch, so you save money at every step of the process.

About Catalyst

Catalyst designs, builds and hosts the world’s fastest and most powerful document repositories for large-scale discovery and regulatory compliance. We back our technology with a highly skilled Professional Services team and a global partner network to ensure the best e-discovery experience possible.
Catalyst Repository Systems

1860 Blake Street, 7th Floor
Denver, CO 80202

Phone: 303.824.0900 | Toll Free: 877.557.4273
Fax: 303.293.9073 |  Privacy PolicyPrivacy Shield


We are excited for our 10:30 am session today "Practical Strategies for Discovery Budgeting and Cost Control" in se…

#clocadoodledoo from the show floor! We are excited for the first full day of @cloc_org. Come stop by booth 151 and…

RT @cloc_org: Bags packed, app downloaded, networking plan in place? Check, check, CHECK! It's a go for #CLOC2018! See you all at the Bella…