by Mark Ettinger – Linguistic Systems, Inc.
on August 28, 2019
Review & Production
The primary aim of e-discovery software is to process and analyze tens of thousands of documents (sometimes millions) in the quest to find and organize relevant files with incredible speed—while understanding the meaning and subplots within. When it’s done well, these platforms save tremendous amounts of time and money.
Still, e-discovery can grind to a halt—or go terribly wrong—if the context of foreign language content is misunderstood. It takes a combination of human and artificial intelligence to get that context right.
Producing correct meaning is one of the biggest challenges for e-discovery translation software. In fact, this is a challenge for any translation project. Getting the literal definitions of words is not enough; meaning can change greatly depending on how words are used in a sentence.
Fortunately, for translating foreign language documents there is a powerful tool—the custom bilingual glossary. It exists to clarify context for important yet still nonsensical or ambiguous passages.
Machine translation is a helpful way for legal teams to get a quick understanding of the foreign language content in their e-discovery databases. But you have to be careful with raw machine translation output.
Why is that? Context is often misunderstood or misrepresented by machine translation workflows. Here are some of the main reasons.
Combinations of these technical and linguistic complications mean that automatically translated data just won’t cut it when it’s at the crux of your case. So how can you bridge the gaps in understanding your foreign language data without breaking the bank?
Custom bilingual glossaries are essentially a list of words and expressions that contain either “do not translate” guidance or precise instructions on which meaning to use in a particular context. Think of the glossary as the specialized code to translate predetermined words in a specific language pair, and for a specific client and legal case.
Custom bilingual glossaries are created by skilled translators and terminologists working in cooperation with the legal team and/or client representatives. After the first glossary iteration, the team typically reviews subsequent passes of machine translation output to identify additional words or expressions that have been mistranslated by the software.
The bilingual glossary is created to add preferential information that instructs the engine to make correct choices in accordance with the needs of the case. Several iterations are usually necessary to fully obtain the desired results. Since the changes involve key passages of the text, the improvement in output performance is sometimes amazing. You get better readability and accuracy.
Custom bilingual glossaries are a small investment, yet they deliver great value in improving accuracy and comprehension. By working with a trusted, experienced language services partner to build a custom language glossary for your next e-discovery project, you’ll be producing a higher quality output that gives your team a better chance of securing a positive case outcome.
e-Discovery is still a significant investment, and there are many checks and balances in place to ensure you strike the right note when it comes to comprehension and cost. Custom bilingual glossaries can help you find that equilibrium.
Mark Ettinger is vice president of Linguistic Systems, Inc. in Boston, MA. Linguistic Systems offers more than 50 years of translation services experience across multiple industries. Ai Translate by LSI is the company’s technology-based solution that has served the legal industry and e-discovery service providers for nearly a decade. The Ai Translate Plugin for Relativity is current through all supported versions of Relativity and RelativityOne. For more information, visit us here and here.
Hacks for Finding Foreign Languages in Your e-Discovery Data
The Search Technique You Should Be Using (But Probably Aren't)
A Beginner's Guide to Service Options in e-Discovery Translation