Abracadabra.
Wouldn’t it be nice to have a computer program that would instantly and magically translate your documents into your customers’ various languages? Click an icon, and presto!
After all, there are Machine Translation (MT) systems out there. Prices range from millions of dollars to free (Google). Just think, you wouldn’t need translation companies, such as One Planet, whose human translators translate merely 2,000 words per day.
Still working the kinks out.
Can computers really do a better job than human translators? The truth is, with the current state of natural language processing, science cannot deliver an efficient MT system of acceptable quality.
To begin with, sophisticated MT systems require elaborate pre-translation programming by specially trained linguists. Depending on the algorithms applied, some terminology updates can take longer than 15 minutes per single term.
After the MT process is over, it still must be proofread by a native speaker. Why? We humans solve many semantic ambiguities without even realizing it.
Ambiguities aside, MT is incapable of solving several other problems related to cognitive and morphological differences in the way we structure our thoughts within various cultures.
When one considers the amount of time and the level of expertise MT requires it becomes a proverbial Stone Soup.
That is why Caterpillar, having invested $25 million into a proprietary MT system, is not using it today.
To better understand problems associated with MT contemplate the following three different meanings of the word “paper.”
Your paper was very good.
I think there will be a story about it in tomorrow’s paper.
Please print it on a high-grade paper.
Modern computer technology is unable to determine and address ambiguity in meaning.
Fitting a square peg into a round hole
Not all applications of MT are doomed to failure. It can be cost-effective when the approximate gist of the content is deemed sufficient. Frequently, MT systems can prove useful for internal communications.
We used MT to help DuPont’s licensing department sort through thousands of foreign patents. A rough translation was suitable for patent-search specialists to uncover any possible patent infringements.
Nuclear engineers at Westinghouse Water Reactor division used MT to bulk screen hundreds of pages per day of technical data—with keywords highlighting problems.
If you think that an MT system could benefit your company, talk to us. We can save you substantial time and money setting it up.
Stone Soup
Long ago, somewhere in Eastern Europe, there was a great famine in which people jealously hoarded whatever food they could find—hiding it even from their friends and family. One day a soldier came into a village and began asking questions, as if he planned to stay overnight.
“There’s not a bite to eat in the whole province,” he was told. “Better keep moving on.” “Oh, I have everything I need,” he informed the villagers. “In fact, I was thinking of making a delicious stone soup to share with all of you.”
The soldier removed an iron cauldron from his wagon, filled it with water and built a fire underneath it. Then, with great reverence, he drew out an ordinary-looking stone from a velvet bag and dropped it into the water.
By now, hearing the rumor of food, most of the villagers had come to the square. As the soldier sniffed the “broth” and licked his lips in anticipation, hunger began to replace the villagers’ skepticism.
“Ahh,” the soldier said to himself rather loudly, “I do like a tasty stone soup. Of course, stone soup with cabbage — now that’s hard to beat.”
Soon a villager hesitantly approached holding a cabbage he’d retrieved from its hiding place and added it to the pot. “Capital!” cried the soldier. “You know, I once had stone soup with cabbage and a bit of salt beef and it was fit for a king.”
The village butcher supplied some salt beef.
And so it went, other villagers brought potatoes, onions, carrots, mushrooms and other ingredients until there was indeed a delicious meal for all. After they feasted, the villagers offered the soldier a great deal of money for his magical stone, but he refused to sell and traveled on.
The future is here to stay
MT is evolving. Simple MT programs available on the Internet can provide some translation capability, mostly through simple systems like Direct and Transfer batch processing.
These “instant translation programs” are not yet capable of doing business translations.
Their popularity however, demonstrates the acute need for on-demand translation.
Scientists around the world are working on natural language processing solutions: the meaning of words, sentence generation, semantics and context.
We hope that automated on-demand translation will soon become a reality. It will be an integral part of electronic publishing systems as well as your globalization management workflow environment.
Polyglot wizards would instantly translate your e-mails from multiple languages and then would deliver your e-mails in a crisp and clear target language of your choice.
When that day comes, One Planet will be the first to embrace it. We have always been at the forefront of new technology and we always will be. Stick with us. The future is here to stay.