Crowdsourcing and machine translation: the start of a beautiful friendship
Machine-aided translation is one of those things people love to hate. Despite the best efforts of enthusiasts like myself, the majority of computer users still believe that machines are useless translators.
The whole area of machine translation has a terrible image problem. There are endless jokes and “true” stories about computer translation failures. Some of these are very funny (like the machine that apparently translated the English saying “out of sight, out of mind” into “invisible idiot” in Russian). However with a little crowdsourcing help, I suspect the machines may have the last laugh. (more…)
Google discontinues its Translation API
Google recently took the decision to retire its widely adopted API, stating “substantial economic burden caused by extensive abuse” as the reason.
The API has been “officially deprecated” since the 26th of May, and will cease to exist completely this December the 1st.
New Translation App Released
A new app has recently been released called “Babelshot (photo translator)”. The clue is in the name – it is indeed an app that will translate the text that appears in a photo taken with your camera phone or that is manually typed into the app.
The app provides translation in 56 languages, from Afrikaans to Welsh to Korean to Slovak, so there aren’t too many countries in which you would be left stranded without help. Photos that the app can translate can vary from signs, to menus, to newspaper articles, and can be particularly useful when travelling abroad in a country where you don’t speak the language. In some cases, it may be vital that a sign can be understood: “no swimming”, “No entry” or the slightly more extreme “Do not enter, danger of death” are quite important!
This new app allows you to take a photo of a portion of text, send it using the app, and receive the translation. The only necessity, other than a desire to know what something means, is a connection to the internet.
The app uses Google Translate in order to provide an instantaneous response. Ah, Google Translate… once again, we return to the topic of machine translation and its reliability. Whilst a translation app for phones is no doubt a fantastic idea, given that you can easily carry your mobile with you wherever you go (and a lot of people already do), and they are, generally speaking, a lot smaller and lighter than a dictionary, can you rely on the translation you’re given?
At the time of writing, there are three users who are not enamoured with the new app: the first three reviews on the Babelshot app page on the Apple website are less than complementary. Reviewr100, for example, had problems using the app when taking a photo of white text on a dark background. A developer promised to fix the issue. However, the user had no more success with black text on a clean white page. Their final comment? “I guess this app may work if you only use it under pristine laboratory conditions and not in the real world.” Another reviewer was compelled to write a comment, despite never usually doing so, and comments that it just doesn’t work, plain and simple. The third review, I must admit, is my favourite. Rearend’s comment, with a title of “Horrible”, simply states: “Doesn’t work-waste of $-junk”.
Obviously being able to translate foreign texts, in whatever form, quickly and easily can be very useful when abroad. I have never used this app, and would be interested to hear any reviews, positive or negative, from those who have – is the Apple site‘s description correct? According to them: “Take a photo of a text, a sign, a book, a newspaper… and Photo Translate will recognize the text and translate it automatically to the language of your choice”? Or is it, as Rearend comments, a waste of money?
Google to machine translate patents
Google has confirmed that it will machine translate patents into more than 29 languages, using the Google Translate interface.
On 30th November, an agreement was reached between Google and the European Patent Office (EPO), in order to facilitate the understanding of patents throughout the world.