Google's Sensitive Translation Service

Eric Baković from Language Log noticed a subtle feature of Google Translate. Google's machine translation system shows radically different results when you change the punctuation or the case of a text.

Here's an example of a small change in a text that improves Google's translation:

[Spanish] tu hija que te quiso tanto y no supo demostrarlo - perdoname.
[English] your daughter that you loved so much and she could not prove it - pardon me.

[Spanish] Tu hija que te quiso tanto y no supo demostrarlo - perdoname.
[English] Your daughter who loved you so much and failed to prove it - pardon me.

"I don't pretend to know anything about Google's translation algorithm(s), but I do find it interesting that what seem like very minor manipulations like those shown above can lead to both bizarrely different results as well as to subtle improvements," notices Eric.

Jim Regan offers a possible explanation: "Google uses statistical machine translation, so algorithms have little to do with it - the translation is created by matching all the translations available for the different parts of the sentence, and then ranked against an n-gram language model of the target language to see how likely it is that those particular phrases go together, to assemble the translation. As case can be significant - acronyms are usually all upper case, proper names use an initial capital, etc. - it makes sense that it affects the translation."

Labels

Web Search Gmail Google Docs Mobile YouTube Google Maps Google Chrome User interface Tips iGoogle Social Google Reader Traffic Making Devices cpp programming Ads Image Search Google Calendar tips dan trik Google Video Google Translate web programming Picasa Web Albums Blogger Google News Google Earth Yahoo Android Google Talk Google Plus Greasemonkey Security software download info Firefox extensions Google Toolbar Software OneBox Google Apps Google Suggest SEO Traffic tips Book Search API Acquisitions InOut Visualization Web Design Method for Getting Ultimate Traffic Webmasters Google Desktop How to Blogging Music Nostalgia orkut Google Chrome OS Google Contacts Google Notebook SQL programming Google Local Make Money Windows Live GDrive Google Gears April Fools Day Google Analytics Google Co-op visual basic Knowledge java programming Google Checkout Google Instant Google Bookmarks Google Phone Google Trends Web History mp3 download Easter Egg Google Profiles Blog Search Google Buzz Google Services Site Map for Ur Site game download games trick Google Pack Spam cerita hidup Picasa Product's Marketing Universal Search FeedBurner Google Groups Month in review Twitter Traffic AJAX Search Google Dictionary Google Sites Google Update Page Creator Game Google Finance Google Goggles Google Music file download Annoyances Froogle Google Base Google Latitude Google Voice Google Wave Google Health Google Scholar PlusBox SearchMash teknologi unik video download windows Facebook Traffic Social Media Marketing Yahoo Pipes Google Play Google Promos Google TV SketchUp WEB Domain WWW World Wide Service chord Improve Adsence Earning jurnalistik sistem operasi AdWords Traffic App Designing Tips and Tricks WEB Hosting linux How to Get Hosting Linux Kernel WEB Errors Writing Content award business communication ubuntu unik