Feeds

How Google translates without understanding

Most of the right words, in mostly the right order

  • alert
  • submit to reddit

Secure remote control for conventional and virtual desktops

Column After just a couple years of practice, Google can claim to produce the best computer-generated language translations in the world - in languages their boffin creators don't even understand.

Last summer, Google took top honors at a bake-off competition sponsored by the American agency NIST between machine-translation engines, besting IBM in English-Arabic and English-Chinese. The crazy part is that no one on the Google team even understands those languages.... the automatic-translation engines they constructed triumphed by sheer brute-force statistical extrapolation rather than "understanding".

I spoke with Franz Och, Google's enthusiastic machine-translation guru, about this unusual new approach.

Sixty years of failure

Ever since the the Second World War there have been two competing approaches to automatic translation: expert rules vs. statistical deciphering.

Expert-rule buffs have tried to automate the grammar-school approach of diagramming sentences (using modifiers, phrases, and clauses): for example, "I visited (the house next to (the park) )." But like other optimistic software efforts, the exact rules foundered on the ambiguities of real human languages. (Think not? Try explaining this sentence: "Time flies like an arrow, but fruit flies like a banana.")

The competing statistical approach began with cryptography: treat the second language as an unknown code, and use statistical cues to find a mathematical formula to decode it, like the Allies did with Hitler's famous Enigma code. While those early "decipering" efforts foundered on a lack of computing power, they have been resurrected in the "Statistical Machine Translation" approach used by Google, which eschews strict rules in favor of noticing the statistical correlations between "white house" and "casa blanca." Statistics deals with ambiguity better than rules do, it turns out.

Under Google's hood

The Google approach is a lesson in practical software development: try things and see what sticks. It has just a few major steps:

  1. Google starts with lots and lots of paired-example texts, like formal documents from the United Nations, in which identical content is expertly translated into many different languages. With these documents they can discover that "white house" tends to co-occur with "casa blanca," so that the next time they have to translate a text containing "white house" they will tend to use "casa blanca" in the output.
  2. They have even more untranslated text in each language, which lets them make models of "well-formed" sentence fragments (for example, preferring "white house" to "house white"). So the raw output from the first translation step can be further massaged into (statistically) nicer-sounding text.
  3. Their key for improving the system - and winning competitions - is an automated performance metric, which assigns a translation quality number to each translation attempt. More on this fatally weak link below.

This game needs loads of computational horsepower for learning and testing, and a software architecture which lets Google tweak code and parameters to improve upon its previous score. So given these ingredients, Google's machine-translation strategy should be familiar to any software engineer: load the statistics, translate the examples, evaluate the translations, twiddle the system parameters, and repeat.

What is clearly missing from this approach is any form of "understanding". The machine has no idea that "walk" is an action using "feet," except when its statistics tell it the text strings "walk" and "feet" sometimes show up together. Nor does it know the subtle differences between "to boycott" and "not to attend." Och emphasized that the system does not even represent nouns, verbs, modifiers, or any of the grammatical building blocks we think of as language. In fact, he says, "linguists think our structures are weird" - but he demurred on actually describing them. His machine contains only statistical correlations and relationships, no more or less than "what is in the data." Each word and phrase in the source votes for various phrases in the output, and the final result is a kind of tallying of those myriad votes.

Security for virtualized datacentres

More from The Register

next story
Microsoft WINDOWS 10: Seven ATE Nine. Or Eight did really
Windows NEIN skipped, tech preview due out on Wednesday
Business is back, baby! Hasta la VISTA, Win 8... Oh, yeah, Windows 9
Forget touchscreen millennials, Microsoft goes for mouse crowd
Apple: SO sorry for the iOS 8.0.1 UPDATE BUNGLE HORROR
Apple kills 'upgrade'. Hey, Microsoft. You sure you want to be like these guys?
ARM gives Internet of Things a piece of its mind – the Cortex-M7
32-bit core packs some DSP for VIP IoT CPU LOL
Microsoft on the Threshold of a new name for Windows next week
Rebranded OS reportedly set to be flung open by Redmond
Lotus Notes inventor Ozzie invents app to talk to people on your phone
Imagine that. Startup floats with voice collab app for Win iPhone
'Google is NOT the gatekeeper to the web, as some claim'
Plus: 'Pretty sure iOS 8.0.2 will just turn the iPhone into a fax machine'
prev story

Whitepapers

Forging a new future with identity relationship management
Learn about ForgeRock's next generation IRM platform and how it is designed to empower CEOS's and enterprises to engage with consumers.
Storage capacity and performance optimization at Mizuno USA
Mizuno USA turn to Tegile storage technology to solve both their SAN and backup issues.
The next step in data security
With recent increased privacy concerns and computers becoming more powerful, the chance of hackers being able to crack smaller-sized RSA keys increases.
Security for virtualized datacentres
Legacy security solutions are inefficient due to the architectural differences between physical and virtual environments.
A strategic approach to identity relationship management
ForgeRock commissioned Forrester to evaluate companies’ IAM practices and requirements when it comes to customer-facing scenarios versus employee-facing ones.