Document Translation

Machine Translation Gaining Speed

July 25, 2013
Technical computer screen

Technical computer screen

Rob Vandenberg, CEO of Lingotek, a crowdsourced-based language translation company, says people expect translations to happen at the speed and accuracy of a Star Trek Universal Translator. Unfortunately, that time is not here yet, but we may be approaching it quickly.

On the CMS Wire, Vandenberg says it would take an army of machine translators an estimated 1,000,000,000,000,000 finely crafted translations to statistically gain the accuracy rate of a real human translator. Big number, yes, but he says it’s doable.

Overcoming Machine Translation Issues

More accurate machine translations can be accomplished if we produce translations more efficiently in a cloud-based collaborative and accessible way. Here’s Vandenberg’s reasoning:

Today there is no consistency with the way translation memory is generated and stored. Common approaches allow analysts to align previously translated documents in order to generate translation memories for the potential benefit to a Computer Aided Translation (CAT) tool in the future. Vandenberg says that these tactics waste time and provide very little gains down the road, unless all content is domain specific.

Machines Can’t Communicate As Well As Linguists

Machine translation usage today is also a sticking point when generating content for end-users. Machine translation engines don’t get the message across to the customers nearly as well as a degreed linguist trained in the culture and nuances of the target language.

But things are rapidly changing. Despite the fact that the Internet doubles in capacity every 18 months and has access to the largest collection of information known to man, it’s still in its adolescence. Machine translators on the Internet will present their best possible guesses at how to translate something, based on a large statistical set that would make the average person’s head spin, but the Internet needs the help of professional and novice translators to help the machines learn.

To Advance Machine Translation Break Content Into Segments

Content entering the cloud-based system should be broken up into small, manageable parts called segments. These segments make it easier for the computer to make its best guess and to receive help along the way. The segment is first populated with the computer’s estimated translation. This typically results in a translation that’s close, but the translation still needs work to communicate the entire message clearly.

Translation machines could analyze each segment to see if it has been translated before. Sources can include a client’s aligned documents or thousands of previous translations. The system looks for exact matches and swaps them with the computer’s best guess. In most cases, the document is now 65 percent accurate and conveys the information intended for a client’s audience.

Still No Substitute For Bilingual Speakers

Human translators are then needed to fill in the missing 35 percent. There is no substitute for bilingual speakers who have both a good handle on the target language and the company’s messaging in mind when they translate. And human translators can add knowledge of the culture and the slang in any given language.

Like Vandenberg, translator Rose Newell thinks a universal translator may one day be built. However, she thinks the ultimate translation machine will have to become more human and less machine-like. “This machine must be able to interact with our society, understand jokes, think not just objectively but subjectively, creatively, and even sensitively to the character of its audience. It must think like a human,” Rose says.

She adds that problems will always arise when profiteers seek to separate translation from its intrinsic human element. She writes that machine translation will never succeed as long as there is no natural language understanding. The ultimate translation machine must understand the intricacies of meaning, grammar, dialect, emphasis, errors and cultural references in a text. Without that ability, machine translation will remain unable to produce reliable translations.

What do you think? Will machine translation continue getting better and better? Will machines one day replace human translators? Send us your thoughts.


About Interpreters and Translators, Inc.

iTi’s dedicated and experienced team offers a wide range of multilingual solutions for domestic and global corporations in a variety of industries. Do you require translation services to enhance your global marketing and sales initiatives or interpreter services to communicate across languages? We specialize in custom language solutions and work with over 200 languages so regardless of the barrier you face, we will work together in synergy to bridge the gap to ensure success. Please feel free to contact us through a message or by calling 860-362-0812. Our offices are open 24/7/365 so we can respond immediately to your interpreting or translation needs anytime, anywhere.

You Might Also Like

No Comments

Leave a Reply