Machine Translation: A Cost-Effective Solution for the eDiscovery Toolbox
With the creation of the Internet and instant communication, global business operations are now the norm in today’s society. Corporations have business deals or offices all over the world. Therefore, it is now commonplace to have documents within electronic discovery collections that contain foreign languages, which can be an obstacle for English-speaking legal teams. Handling documents in foreign languages has become much simpler and more cost-effective than in years past, and legal teams should understand their options.
The standard for document translation for many years has been human translation. Documents are provided to a company or the case team hires a document translator to read the documents and write out the translation. The main benefit of this type of translation is that the translation is extremely accurate and written in true sentence form. Different languages use different sentence structures. A sentence translated from Spanish to English without reconfiguring may not make complete sense. The human translation method allows the translator to not only convey the words but to set up the sentences to match the syntax used for the particular target language. The main problem with human translation is cost. Human translation is expensive and can be cost-prohibitive for anything other than a few key documents, which is an issue, as the team will rarely know if a document is important until it has been translated.
Technology improvements over the last few years have allowed machine translation to become more prevalent. There are many different types of machine translation tools, each of which typically involves uploading the text of documents to a server housing the translation software, which will then apply a combination of a computer algorithm and a word dictionary to translate the documents from one natural language to another. The primary benefit of this type of translation is cost. Machine translation is incredibly inexpensive and affordable for large amounts of documents. The other benefit to machine translation is speed. In many cases documents can be translated in a couple of hours. The biggest issue with machine translation is that the machine is translating a word in the foreign language to its counterpart word in the target language, but not making any adjustments to the sentence structure. This can leave the sentence and its meaning hard to understand due to the different placement of the wording in the target language. The other issue is that five words in Spanish may all translate to one word in English. This can lead to confusion and at times cause the case team to incorrectly interpret what the author was stating.
The recommendation when it comes to translation is to use a combined approach. The cost of machine translation allows for large amounts of documents to be dealt with at one time and gives the case team most of what they need. Key documents or documents where the meaning is not quite clear can be uploaded to a human translator to retranslate using the proper sentence structure. Even though some documents are being processed twice, the savings can be quite large due to the great difference in pricing between machine and human translation. Many translation service vendors also offer a combined translation service. The document will be machine translated and a reviewer will then give it a quick glance and make corrections as necessary. This is a little bit more expensive than straight machine translation, but is another cost-effective tool in the arsenal.
The ability to translate documents is something that all case teams need to have in their toolbox. As business becomes increasingly global in nature, electronic discovery collections will increasingly contain documents in many different languages. Make sure that your team is ready to put the right tools to work!