Skip to content

GTS Blog

by Dave Grunwald, CEO of gts-translation.com

Archive

Tag: Machine Translation

Anyone remember the Y2K scare? For months leading up to the new millennium, the news media brought us lurid predictions of the havoc which would be wreaked by the Y2K effect. Would planes fall out of the sky? Would the pumps at the gas stations stop working? The power grid? Would the bank computers know how much money we have? Lots of people, myself included, were worried and waited in anticipation for the clock to strike midnight on 1.1.2000.

Why am I bringing this up? Well tonight may be doomsday for developers who built applications around the Google Translate API, and users of those applications. In it’s announcement last May, Google announced that it would be terminating its free Translate API service.

 The courtesy limit for existing Translate API v2 projects created prior to August 24, 2011 will be reduced to zero on December 1, 2011. In addition, the number of requests your application can make per day will be limited. Google Translate API v1 will be shut off completely on the same date (December 1, 2011).

So tonight is the night. Just like in the story of Cinderella, the fair maiden must leave by midnight or be revealed as a chambermaid. But will she leave a glass slipper behind? Will the prince be able to find the fair maiden with the only foot to fit the slipper?

What will happen to all of the applications that run on the free Google translations? Will the translations stop working? Will we start to see a Google message that says “this website needs a different google api key” (like it does for Google Maps)? Will the developers who switched to the Microsoft Translator API need to discontinue some of the languages (since Google has some language pairs that Microsoft does not). Will some developers start to charge money for automatic translation services? I’ll be on the lookout for answers to these questions and I hope you will too. Please report anything you see.

Another issue I am interested in is security. Now that Google is charging for use of its API, will hackers try and target the API? A blog post I just saw warns against this very eventuality. Will some clever developer find a way to spam the (still) free Google translation website and mine the translations for free? Will Google be able to clamp down on this? These are questions which will be interesting to learn more about as the events unfold.


In translation industry news: Transperfect is continuing its shopping spree by acquiring Worldlingo, the company that pioneered the Internet translation industry (read the news here).

Worldingo was founded in the late 1990s by Australian Phil Scanlan, an independently wealthy and very shrewd business person. Mr. Scanlan did a few things that caused Worldlingo to prosper. He purchased a Systran Enterprise Server 5 system and started selling machine translation solutions for websites years before anyone had even dreamed about doing it. He gave away free website translation widgets that linked back to the Worldlingo site. This gave Worldlingo number 1 ranking in nearly every keyword phrase associated with translation. And he made a deal with Microsoft that integrated Worldingo into every Microsoft Office product software license so when someone asked for translation from Microsoft Word, for example, they received the free translation from Worldlingo.

All of this gave Worldlingo total domination of the search engines for years. And they made a ton of money. But since then a few things have happened: (1) Phil Scanlan left Worldlingo in pursuit of other endeavors. (2) Microsoft dumped Worldlingo in favor of their own MT which is now integrated in MS Office. (3) Google and Microsoft offer free translation widgets so who needs Worldingo for that? (4) Worldlingo stagnated with outdated Systran 5 technology and left it there for years. Only about 2 years ago did they integrate Language Weaver into their website. (5) Many new companies have started to resell MT solutions and Worldlingo is now just one of several in the field. (6) Worldlingo has been dropping like crazy in the search engine rankings and is continuing to lose ground. And Worldlingo lives on Internet traffic.

Which leads us to Transperfect. What are they buying? MT technology? Worldlingo does not have any proprietary MT technology that I know of and licenses software from MT vendors. Customers? Customer loyalty in the online world is fickle (how can you be loyal to a website?). Worldlingo does have a very snazzy online ordering system. But they now face stiff competition from other online translation companies like Babylon and MyGengo. The software solutions they are selling for website translation are outdated and now face stiff competition from companies like ToLingo and Smartling.

The main asset which Transperfect is buying, in my opinion, is a (still) strong online presence. Something that Transperfect is lacking. The question is: will Transperfect breath new life into Worldlingo? Will they integrate new technology which will revive the website and make it a dominant player in face of all the new competition? Will they pump in funds to improve the site in face of the growing competition in the online translation world? Will they be able to help Worldlingo stop their online decline?

Transperfect’s acquisition policy, to the best of my knowledge, is to let the management continue to do their own thing after the deal has been made. The big question is: will that save Worldlingo?

Microsoft Office 2007 Logo

Image via Wikipedia

For years MS Office applications, like MS Word, have come bundled with spell checkers and grammar checkers in different languages. Everyone knows that. But now it seems that the folks at Redmond have come up with a way to simultaneously spellcheck documents that are written in a combination of different languages. The NY Times featured a blog post about it a few days ago. Click here for the Microsoft web page that explains this exciting new feature in detail.

This feature has huge implications for machine translation systems. One of the problems that many MT users experience is translating web pages or documents that are written in several languages. MT systems go from one language to another. So for example if you have a blog post which has comments in several languages, the MT is unable to translate everything and will just pass through some of the text as-is. The following screenshot says it all:

Now this may all change since one of the biggest MT vendors is Microsoft, the company that develops MS Office. Maybe the MS Research people who run the Bing Translator project will talk to the MS Office people and bring over this great feature. I wrote a blog post last year about integrating spell and grammar checking tools into the MT post editing process with information which is relevant to the transfer of language technology from MS Office to Bing Translator.

Perhaps other MT vendors can implement this feature by combining language recognition and spell/grammar checking. I am sure that the multiple language spellcheck feature will cross-over to MT sometime in the not too distant future.

PS Don’t forget to try our new GTS Website Translator at http://webtranslator.gts-translation.com.

Enhanced by Zemanta
An IKEA Billy bookshelf, 80x106 cm, finished i...

Image via Wikipedia

I saw an interesting article yesterday. It appears that Ikea is predicting the demise of printed books altogether by redesigning its book shelves to be bookless. The bookshelves of the future, according to the Swedish furniture king, will have just about everything else besides books. Books are dying. One of the top book vendors in the country Borders, announced that it is shutting down last July. Book sales are sharply down and the bottom isn’t in sight. Newspapers are also hurting. Newspaper publisher Rupert Murdoch recently closed one of the world’s oldest newspapers.

So what does this mean to the man on the street? What does it mean to the translation industry? Here are some of my own thoughts on this:

  • Shush! Libraries will become obsolete too. How can you have libraries without books?
  • Demand for translation services is going to continue to increase. People will read less but will be spending much more time reading online. The demand for related software products and online content will increase and all of that will need to be localized into multiple languages.
  • Printed manuals like the ones provided with cars and electronics devices will become obsolete in the next few years. All support information will be provided online. This prediction was made by SDL CEO Mark Lancaster at the AMTA conference in Denver last year.
  • The demand for DTP software and services will drop as more and more content will be published using markup languages like XML.
  • Demand for machine translation will increase. So will the quality of MT systems. This will be driven by the explosion of online content that we are seeing in recent years.
  • More and more content will be generated by software, rather than by humans. Semantic web publishing will take off. The prose of this content and its controlled nature will lend itself to machine translation much better than the printed word of old. MT systems of the future will be capable of publishing content into multiple languages much faster than today.
  • Robots will be trained to write. OK, granted that this it very futuristic but is it so far fetched? Robots will write books, robots will read books. What will humans do? Get coffee for the robots I guess.
Enhanced by Zemanta

The Inside Facebook blog reported last week that Facebook is testing a Translate button for comments on pages. With the introduction of this new feature, comments that are written in languages other than your account’s default language now include a Translate button next to them. When you click on the Translate button, the comment is automatically translated to your default language. The Translate button is then replaced by an Original button which untranslates the comment.

I have learned that this service will be provided through the Microsoft Translation API. Facebook will not be using a self-developed machine translation for this service. At least not at this point.

Microsoft is fast becoming the ‘go-to’ translation API player. Since Google deprecated their free API and turned it into a paid service, developers are abandoning Google and using Bing instead. Microsoft has stressed over the past few months that they will make no changes to their free translation API and have even increased user limits to make it even more attractive to developers.

Another major software company that went with the Bing Translation API is Autodesk, recently releasing TransTips for AutoCAD, Inventor, Revit, and 3ds MaxThis plugin translates the contents of tooltips in various Autodesk products automatically into one of 35+ languages via the Bing translation API.

The cooperation between Facebook and Microsoft should come as no surprise to anyone. Microsoft is a minor shareholder in Facebook, having bought 1.6% of the company shares for $240 Million back in 2007. I tried to get more information from Microsoft about this new cooperation with Facebook but could not get any information. I can assume with great certainty however, that the terms of the cooperation were discussed at a high level and that Microsoft will provide specially trained MT engines to Facebook at some point in time.

I also tried to obtain more information from Ghassan Haddad, Director of Internationalization at Facebook. I’ll report back with any new information I can obtain.

Enhanced by Zemanta