Machine Translation

The Future of Translation

June 22, 2011/0 Comments/in Language Blog /by Alison Kroulek

Ray Kurzweil is a pioneer in technological fields including speech recognition and text-to-speech synthesis. He also believes one day soon, computers will develop their own consciousness and superhuman levels of intelligence. In fact, in an interview with the Huffington Post, Kurzweil told industry guru Nataly Kelly that he believes computers will be able to translate as well as human translators by 2029.

At the moment, it’s obvious that machine translation has a way to go. But it’s improving all the time,and what if Kurzweil is right? Is there a future for translators in the brave new world he predicts? Fortunately, the answer is yes. Kurzweil says that:

“These technologies don’t replace whole fields, in general. What they do is replace a certain way of applying them.”

The translation field is always changing, but translation companies that are willing to change with the times should have no problem thriving, according to Kurzweil:

“These tools are going to increase our ability to use, create, understand, manipulate and translate language. The idea is not to resist the tools, but to use them to do more.”

Why Machine Translation is Not Good Enough

May 11, 2012/0 Comments/in Language Blog /by Alison Kroulek

Machine translation that’s good enough to substitute for human interpreters is like the great white whale, sought by science fiction writers, businesses and militaries alike. However, despite all the hype about the latest iPhone translation app and the ubiquity of Google Translate, nobody has yet managed to produce an algorithm that does the job as well as a bilingual human.

A recent article in Slate on the efforts of the US military to develop a machine translation device to substitute for human interpreters in Afghanistan is a case in point. The article describes the results of a 5 year research effort funded by DARPA. The snippet below shows just how well the device performed in place of a Pashto interpreter:

Rachel asked: “Would you introduce me to him?” Aziz failed to understand the machine’s translation (though he does speak English), so she asked again: “Could you introduce me to the village elder?” This time, there was success, after a fashion. Aziz, via the device, replied: “Yes, I can introduce myself to you.”

Unfortunately, Aziz was not the village elder in question. C3PO, where are you when we need you? DARPA’s speech-to-speech translation system, called TransTac, achieved an 80% accuracy rate by the end of the research project. Obviously, these are not the droids we were looking for.

The problem, as Slate points out, is that computers are great at storing knowledge and making calculations, but they lack the key ingredient of a successful interpreter: understanding. Attempts to add this essential human ingredient by comparing machine translations to human-created translations and by having real people rate translations for quality also tend to fall short. Plus, it’s slow and expensive. As Slate writer Konstantin Kakaes put it:

The difficulty of knowing if a translation is good is not just a technical one: It’s fundamental. The only durable way to judge the faith of a translation is to decide if meaning was conveyed. If you have an algorithm that can make that judgment, you’ve solved a very hard problem indeed.

We couldn’t agree more, and that’s why it’s so important to use trained human translators and interpreters for important communications.

We recently published a further article regarding the pitfalls of MT, you can view it here:

Go to article

Message in a bottle

November 4, 2009/0 Comments/in Language Blog /by Alison Kroulek

A French love letter was found near Falmouth, Cornwall last week.

The beer bottle was found by Martin Leslie, a coastguard manager, and his family as they walked on Praa Sands, near Falmouth.

The bottle was poking out of the sand; its top was sealed with red candle wax. Inside was a three A4 pages handwritten in French and dated September 28th.

Mr Leslie had a go at translating the letter using the internet but could only decipher words relating to love, death and missing someone.

He assumed that the letter must be a suicide note and handed the letter to Falmouth Coastguards to pass onto their counterparts in France.

According to the Telegraph Mr Leslie said the woman said she and her lover shared magical moments together but that she understood that he had to return to his wife. She finished by saying she hoped to find another man like him with whom she could live a beautiful life.

The letter said ”These magic moments are pure secret. The secret of life and pleasure without limits. In twenty years, it will still be here, the previous moments of happiness, when life will get dreary, we will be able to tap into these memories to remember what it is to live again.”

Mr Leslie plans to keep hold of the letter which is unsigned and has no contact address on it.

Cost of NHS Translation

February 17, 2012/3 Comments/in Language Blog /by Alison Kroulek

The National Health Service is currently catching some flak for the costs of its interpreting and NHS translation services. As the BBC reports, a group called 2020Health made a Freedom of Information request to determine the cost of these services to the NHS, and found that last year’s bill came to over £23m.

Speaking to the BBC, 2020Health CEO Julia Manning condemned the costs as excessive, saying:

“The costs involved are truly staggering in an age of austerity. Urgent action must be taken by trusts to stem the flow of translation costs. The most glaring problem is that NHS trusts translate their own material rather than have access to a central pool of translated documents.”

Most bureaucratic systems can be improved, and the NHS is no exception. The 2020Health report offers some good suggestions for cutting costs. For example, the think tank suggests that the NHS create a centralised database of translated documents, so that different trusts have access to each other’s materials and don’t need to order new translations from scratch to convey information that was already translated into the same language at another location. The group also suggests that NHS trusts use standardized procedures to break down and track where the money being spent on translation is going.

In its report, though, even 2020Health admits that translation services are necessary, saying

In order to avoid cuts to services, the NHS is currently looking to make a 20% savings through service redesign. Whilst it might be tempting to cut costs by simply reducing spend on translation services, this is not necessarily the most effective or economical solution.

Healthcare is one of those situations in which accurate translations are much more than just a nicety. Being able to communicate effectively with patients is absolutely crucial to their health and well-being. That’s why some of the recommendations in the report, while undoubtedly well-intentioned, have the potential to do more harm than good.

For example, the report recommends switching as much as possible from human translators to machine translation services like Google Translate. These tools are great for getting the gist of a document, but they often produce translations that are at best awkward and at worst misleading or incoherent. We’ve reported many times on this blog about the dangers of using Google Translate (other online translation tools are available) so we’d say its dangerous when you’re communicating about something as important as someone’s health.

The report also criticises NHS translations for “perpetuat[ing] a system in which [non-English speaking patients] are ostracised from the majority of the English-speaking public.”

Of course, newcomers should be encouraged to learn English. However, when providing medical care, the primary consideration should be getting the patient the information and the help they need, not “encouraging” them to improve their English skills by conveying information to them in a language they don’t adequately understand.

Machine Translation – A Potted History

May 22, 2017/2 Comments/in Inside Track /by Jonny Henchman

The concept of machine translation has existed for centuries, but it was not until the early 1950s that it began to become a reality. Since then, machine translation has advanced hugely, though it still cannot yet compete with the skill and finesse that a human mind can apply to translating a document.

The birth of machine translation

In 1949, Warren Weaver of the Rockefeller Foundation put together a set of proposals on how to turn the idea of machine translation into reality. He blended information theory, code breaking lessons learned during the Second World War and the principles of natural language to pave the way for machines to translate one language to another.

One of the earliest machine translation successes was the Georgetown-IBM experiment. In 1954, IBM demonstrated at its New York office a machine that could translate Russian sentences into English. Though the machine could only translate 250 words (into 49 sentences), the world was delighted by the idea. Interest in machine translation around the world saw money being poured into this new field of computer science. The Georgetown experiment researchers, bursting with the confidence of their initial success, predicted that machine translation would be mastered within three to five years. Read more

Google Translate Now Translates Latin Ad Libitum

October 11, 2010/3 Comments/in Language Blog /by Alison Kroulek

Last week, Google added another language to its popular Google Translate service, and Latin students everywhere breathed a sigh of relief. Yes, Google Translate now decodes Latin. The announcement came via a blog post written entirely in Latin by engineer Jakob Uszkoreit. Show-offs!

Google expects the Latin version of Google Translate to be quite popular with students who are studying the language, as well as for people studying philosophical and other texts originally written in Latin.

The fact that Latin is a dead language should make Google’s machine translation more accurate, as the company explained in its blog (Latin translation from the Telegraph):

“Unlike any of the other languages Google Translate supports, Latin offers a unique advantage: most of the text that will ever be written in Latin has already been written, and a comparatively large part of it has been translated in to other languages. We use these translations, found in books and on the web, to train our system.”

Microsoft Translator, Now in Welsh

February 27, 2014/0 Comments/in Language Blog /by Alison Kroulek

Good news for the Welsh language this week: Microsoft Translator is now available in Welsh. Microsoft Translator provides a variety of translation tools for individuals and businesses, as well as a machine translation API for developers.

The Welsh version of Microsoft Translator is a joint project between Microsoft and the National Assembly for Wales. It was unveiled to the public during a celebration for International Mother Language Day held last Friday in the Senedd. Read more

Gmail Incorporates Automatic Translation

May 21, 2009/4 Comments/in Language Blog /by Alison Kroulek

Google just unleashed a new feature to help users break through the language barrier. After you enable the feature, you’ll be able to translate any email you receive with a quick click of the mouse. The new Gmail translation feature can translate 41 different languages, including Thai, Estonian and Maltese.

Automatic translation is available to all Gmail users including those who use the email software as part of the apps collaboration and communication suite for organizations. It will help users to communicate better in today’s multilingual world.

The goal of this feature is to make it easier for people who work for international companies to communicate. Theoretically, using the Gmail translation feature you could hold a conversation via email with employees from around the world, and each employee would be able to communicate using his or her native language.

Chris Dawson of ZDNet Education sees another possibility for the new feature: allowing students to have email pen pals who speak other languages.

However, it should be noted that computerized translation is far from perfect. No computer program has yet been invented that can correctly translate 100% of conversations from one language to another, especially when figurative language or colloquial expressions are being used. So, messages translated using Gmail’s translation service may come out sounding a little off when read by a native speaker.

This new tool is useful however, users should be aware that machine translation is not always reliable, even Google themselves have acknowledged that machine translation technology isn’t perfect.

Google do maintain that even if mistakes creep into the text, the recipient will be able to get the gist of the message.

In an article on eWeek.com, Jeff Chin, the Project Manager of Google Translate, said as much in an email interview:

“It can be quite useful in providing the quick gist of a message, especially if you receive a lot of e-mails that aren’t in your native tongue,” he wrote. “If the translation is awkward or not quite right, you can quickly return to the original message by clicking ‘View original message’ link.”

If clear communication is your goal, it is advisable to use a professional translation company who can assist you with your translation needs

Google Penalizes Bad Machine Translation

January 11, 2017/5 Comments/in Language Blog /by Alison Kroulek

& 10 tips for good international SEO

Planning to translate your business website into another language? Free, automatic translation tools like Google Translate might seem tempting, but here’s one more reason to avoid relying on them: Google doesn’t like it. And if Google doesn’t like it, you’d better not do it, at least not if you value your website traffic.

It seems odd that the search engine gods would issue penalties for using Google’s own product, but apparently search engine spammers have been publishing lots of awkward, error-laden machine translated content. To keep their results as accurate as possible, Google classifies automatically translated content as “automatically generated content,” which violates their webmaster guidelines.

That means that poorly translated content could seriously impact your rankings. Also, as Ariel Hochstadt pointed out in Search Engine Land, if you’ve monetized your site using AdSense, your account could be disabled for including “websites with gibberish content that makes no sense or seems auto-generated.”

Ironically, Google itself has started using automatically generated content on its own properties, like the Google Play store. However, as Search Engine Land points out, it appears that Google is using some sort of new and improved Google Translate that’s not available to the general public.

Why not release the latest and greatest Translate tool? Hochstadt speculates:

My best bet is that Google is afraid of mass spamming that could be hard to identify. Nevertheless, if they think it is good enough for them to publish it on their Android and Chrome stores, why wouldn’t they allow others to do the same in Google Translate? Knowing Google, you probably are aware that their rules sometimes oblige us, but don’t apply to those located in Mountain View.

Fair or not, you’re better off using a professional translator, or at the very least having the final product reviewed by someone who is fluent in your target language and able to correct any mistakes. To help you out and keep you the right side of the Google police, we have put together a collection of 10 International SEO (Search Engine Optimisation) tips that you can employ to help boost the performance of your multilingual site. Read more

K International on the Gadget Show

October 27, 2015/0 Comments/in Inside Track, Language Blog /by Jonny Henchman

If you were sitting down to watch the Gadget Show on Channel 5 last night (26th of October 2015), you will have seen a couple of our linguists featured in the show. First up, Monica Bloxam took on presenters, Amy Williams and Jason Bradbury, in a Human Vs. Machine translation challenge. Read more

Posts