Slashdot Mirror
Romancing The Rosetta Stone
Roland Piquepaille writes "Not only this news release from the University of Southern California has a fantastic title, it also has a great content. This story is about one of their scientists, Franz Josef Och, whose software ranks very high among translation systems. "Give me enough parallel data, and you can have a translation system for any two languages in a matter of hours," said Dr. Och, paraphrasing Archimedes. His approach relies on two concepts, gathering huge amounts of data, and applying statistical models to this data. It completely ignores grammar rules and dictionaries. "Och's method uses matched bilingual texts, the computer-encoded equivalents of the famous Rosetta Stone inscriptions. Or, rather, gigabytes and gigabytes of Rosetta Stones." Read my summary for more details."
Since I mistakenly borrowed some undubbed Cowboy BeBop.
Romancing the Rosetta Stone
'Give me enough parallel data, and you can have a translation system in hours'
University of Southern California computer scientist Franz Josef Och echoed one of the most famous boasts in the history of engineering after his software scored highest among 23 Arabic- and Chinese-to-English translatio systems, commercial and experimental, tested in in recently concluded Department of Commerce trials.
"Give me a place to stand on, and I will move the world," said the great Greek scientist Archimedes, after providing a mathematical explanation for the lever.
"Give me enough parallel data, and you can have a translation system for any two languages in a matter of hours," said Dr. Och, a computer scientist in the USC School of Engineering's Information Sciences Institute.
Och spoke after the 2003 Benchmark Tests for machine translation carried out in May and June of this year by the U.S. Commerce Department's National Institute of Standards and Technology.
Och's translations proved best in the 2003 head-to-head tests against 7 Arabic systems (5 research and 2 commercial-off-the-shelf products) and 14 Chinese systems (9 research and 5 off-the-shelf). In the previous, 2002 evaluations they had proved similarly superior.
The researcher discussed his methods at a NIST post-mortem workshop on the benchmarking held July 22-23 at Johns Hopkins University in Baltimore, Maryland.
Och is a standout exponent of a newer method of using computers to translate one language into another that has become more successful in recent years as the ability of computers to handle large bodies of information has grown, and the volume of text and matched translations in digital form has exploded, on (for example) multilingual newspaper or government web sites.
Och's method uses matched bilingual texts, the computer-encoded equivalents of the famous Rosetta Stone inscriptions. Or, rather, gigabytes and gigabytes of Rosetta Stones.
"Our approach uses statistical models to find the most likely translation for a given input," Och explained
"It is quite different from the older, symbolic approaches to machine translation used in most existing commercial systems, which try to encode the grammar and the lexicon of a foreign language in a computer program that analyzes the grammatical structure of the foreign text, and then produces English based on hard rules," he continued.
"Instead of telling the computer how to translate, we let it figure it out by itself. First, we feed the system it with a parallel corpus, that is, a collection of texts in the foreign language and their translations into English.
"The computer uses this information to tune the parameters of a statistical model of the translation process. During the translation of new text, the system tries to find the English sentence that is the most likely translation of the foreign input sentence, based on these statistical models."
This method ignores, or rather rolls over, explicit grammatical rules and even traditional dictionary lists of vocabulary in favor of letting the computer itself find matchup patterns between a given Chinese or Arabic (or any other language) texts and English translations.
Such abilities have grown, as computers have improved, by enabling them to move from using individual words as the basic unit to using groups of words -- phrases.
Different human translators' versions of the same text will often vary considerably. Another key improvement has been the use of multiple English human translations to allow the computer to more freely and widely check its rendering by a scoring system.
This not coincidentally allows researchers to quantitatively measure improvement in translation on a sensitive and useful scale.
The original work along these lines dates back to the late 1980s and early 1990s and was done by Peter F. Brown and his colleagues at IBM's Watson Research Center.
Much of the improvement and
when it's in the form of a fish, and can fit in my ear...
"I turn away with fright and horror from the lamentable evil of functions which do not have derivatives."
This could be an amazing improvent to search engines. If they could instantly translate a page before showing it in the results.
This is exactly NOT a universal translator as it uses matched bilingual texts. You need an already translated text for his system to work.
We'll have a supercharged Babelfish ?
You know, it's not really a summary when you just delete half the article.
That reference to DARPA has me a little worried about the sort of uses this technology will be put to. I wonder, are the CIA trying to shore up holes in their translation abilities (particularly for Arabic/etc) by using software. What happens when you pair this technology up with the Echelon project? Are we going to see a dramatic rise in the ability of the government to spy on nationals and particularly foreign nationals now?
All those moments will be lost in time, like tears in rain.
No really... what if it used a shared database and there were hundreds, or thousands, of the systems around the world... Seems like it could become a pretty sophisticated system. And maybe one day it will be available in the form of a small fish which you place in your ear?
Visualize the world of wine
The uber-geeks are going to have a field day with Klingon...
This is a bit of a worry for privacy concerns, given that if I want to keep something secret from the world and private just between me and my intended recipient I have one less option.
How long until this is able to decode things like speech, too, and convert it into something recognisable in another langauge? would it still hold my voice patterns and sound like me? and if it were converted back to the English I already do speak, with mistakes, could that then be used against me in a court of law?
Scary stuff
"Give me enough" is a key element of the Law of Eventuality. Give me enough money, and I'll solve the Microsoft monopoly threat with a hostile takeover. Give me enough time and I'll clean up almost any unnatural disaster site by leveraging nature's own methods.
Give me enough simulated neurons and enough truisms and I'll make a sentient machine.
Eventually, with enough resources, anything is possible. Throwing more time and resources to a problem is rarely exciting science. Reducing the inconveniently large values of 'eventually' and 'enough' are the real problem.
[
'Almost everyone'? What *are* you talking about? You must be an American. From a recent online Harris poll, most Americans think at least half the world speaks English. This is just plain wrong. The truth of the matter is that it's more like 20%. That's it. Most people on the NET might speak English, but most people in the world? Hardly.
My journal has hot
English may be the closest thing we have to a universally-spoken language, but it certainly isn't going to become the -only- language any time soon, if ever. If all other languages disappeared, though, we would definitely need translation for all the literature we have that isn't written in English.
Here is Japanese Slashdot, and I'm sure there are others.
Am I the only one who thinks that translation is quickly becoming obsolete?
Yes.
That's an example from a few years' back of an attempt to translate "the spirit is willing but the flesh is weak" from English to Russian and back to English using a different translator.
Can anyone try this on the new (or some other recent) algorithm?
BTW here's Doc Och's most recent website:
Franz Josef Och
Esteem isn't a zero sum game
I believe that using a statistical approach like this is a step in the right direction. Manually building sets of rules, dictionaries, etc., is a waste of time and hard to do. And manuall-built systems become stale as languages evolve, unless a lot of continuing work is done.
For me the holy grail is when I can converse with a computer meaningfully. I believe a similar approach will be required for the computer to "understand" language, and to be able to formulate a coherent and appropriate response.
If email, IRC/"chat rooms" *spit* and SMS are anything to go by, a great number of young and not-so-young people who *do* have English as a first language are barely capable of forming even simple sentences in it correctly.
Regards,
Tim. (Grumpy old man day)
DARPA actually proposed that a forced conversion to English policy would be more cost effective for the defense department to implement through military invasion than some complicated translation scheme. Hence congress's support for the translation project.
Isn't the Doc supposed to be in the next Spiderman movie?
Universal translator anyone?
Er, aging geek embarrasing self again, mutter...
From the article: his software scored highest among 23 Arabic- and Chinese-to-English translatio systems
Oops - guess we need some more parallel data (or a few more gigs of rosetta stones).
A man who speaks three languages is trilingual.
A man who speaks two languages is bilingual.
A man who speaks one language is American.
DG
Want to learn about race cars? Read my Book
The battle for the Rosetta Stone "Things are looking decidedly rocky at the British Museum - Egypt's leading archaeologist has demanded the return of the Rosetta Stone. But the museum argues that the removal of the four-foot slab that unlocked the mysteries of the pharaohs would be disastrous"
"Read my summary for more details."
:)
I'd rather have less detail in a summary - thanks
English --> French --> English --> Korean --> English. Of course, it helps that the first sentence is munged anyway ;-)
When I am king, you will be first against the wall.
Sounds like a brilliant idea. Hopefully this is something that could eventually be compacted enough to fit into consumer electronics. It would be great to be able to watch TV from every country without any language barrier!
DeviantArt Page
NSFW"Almost anyone"? Show me the data. And of course, you must remember that while many can speak English, fewer can speak it well.
Firstly we could consider the enormous body of work currently available in other languages.
Having this able to be translated into English or other languages could be very valuable for scholars.
Secondly, English is not the primary tongue for the majority of people on the planet - to suggest that because a lot of people can manage to converse in it that the ability to translate between other languages isn't valuable is foolish.
Also note that the article specifically mentions Arabic and Chinese, which I don't think crossed your mind. China has the largest population on the planet, remember.
Translation is far from obsolete, especially given that the majority of the Western world, and especially America, is piss poor at being bilingual.
Star Trek's had a universal translator for years...
Heh. A friend of mine from India who worked for GM was fond of that joke. :)
My journal has hot
How can this system compensate for the different dialects of all of the different languages?
"Some fight for law. Some fight for justice. What will you fight for? One day, you will see."
> Americans think at least half the world speaks English.
Better-informed Americans (a small miniority of the class) would be aware that Spanish is well on the way to becoming the predominant language in the USA.
But, IMHO, English could become the next Latin: the dead language that everybody has to learn if they're going to try and influence the world.
BTW, every "% of humanity" statistic has to consider that most humans are Chinese.
Am I the only one who thinks that translation is quickly becoming obsolete?
Almost everyone can speak, read and write at least tolerable english and most young people can have full fledged discussions in it.
That isn't much help if you want to read (say) De Uitvreter by Nescio and you don't know Dutch, does it? Or for a slightly more geeky angle, if you want to read Edsger Dijkstra's Dutch texts?
JP
Not only it is not obselete in the real world, it can have tremendous influence on how we can decrypt information, when it's a mix of cipher / steganography.
And if it wasn't for Spanish (and South America), Americans would think 100% of the world speaks English..
What is the novelty of this?
It's hardly news that you can always find correlations in two sufficiently large sets of data.
Reminds me of the Steve Martin joke:
"Chicks go for the intellectual types. I figured the best way to impress 'em was to read a lot of books. But hey, do you know how many books there are? Why, there must be, hundreds of them. But I was already a pretty smart guy. I didn't waste my time reading all those books. Heck no.
I read, the dictionary. Hey--I figure it's got all the other books in it."
http://www.britishcouncil.org/english/engfaqs.htm# howmany
Translators are needed for 3/4ths of the world. Not what I would call close to obsolete.
The surprise isn't how often we make bad choices; the surprise is how seldom they defeat us.
Most Americans think at least half the world that matters speaks English.
American attitude is if you don't speak English, you don't matter :)
I have said it before, on
A good translation is based on several non-quantifiable parameters:
Example:
"My controller has failed. He is going to be replaced" can mean:
OK, maybe the above example is not perfect, but you get my drift... Machine Translation? Bah! Humbug.
That was my "machine translation" rant/flamefest of the month. Carry On.
The right to offend is far more important than the right not to be offended. (Rowan Atkinson)
"One of the great advantages of the statistical approach," Och explained, "is that most of the work goes into components that are language-independent. As long as you give me enough parallel data to train the system on, you can have a new system in a matter of days, if not hours."
This statistical method is probably the best approach to computerized translation. It seems to approximate how the human mind will translate a give sentence most efficiently. Language can get awfully complex, and individual words often have, at best, an ambiguous meaning when interpreted alone. One must take into account the context of that word to specify and refine its meaning. This obviously leads to a huge number of permutations to represent a huge variety of thoughts, but the relative size of this number is diminishing as computers become more powerful.
Therefore, instead of playing with messy grammars and sentence structures, we can simply have a catalogue of thoughts as represented by words, and correlate that catalogue with a different set of words to facilitate translation. This software would operate on a deeper level than it would if it operated with the words and symbols themselves. It would utilize a map of the deep structures of language, instead of a map of the less-meaningful words and grammars.
I really like this method, and while it may seem like a brute-force hack applied to translation, the simple fact that languages do not contain elegant patterns must be accepted. It also appears to be a most efficient method, as the simple comparisons involved would bring the speed of translation into realtime.
University of Southern California computer scientist Franz Josef Och echoed one of the most famous boasts in the history of engineering after his software scored highest among 23 Arabic- and Chinese-to-English translatio systems, commercial and experimental, tested in in recently concluded Department of Commerce trials.
Maybe what Dr. Och should do next is write some software to double-check the work of whoever translates his press releases from the original Latin. The translator seems to have missed a few words here and there.
I'm not sure this is really applicable to translating literary works. These kinds of translations require an understanding of the native culture of both the source and target languages, as well as the intent of the writer, in order to generate an understandable translation that the target group can appreciate. A computer translation system like this one is incapable of performing these sorts of analysis.
What this is really good for is on-the-fly translation of material where the reader simply wants to comprehend what was written (think the old babelfish engine). This has obvious applications on the web, as well as many other areas (on-the-fly server-side translation for IM systems, etc, etc).
English : correctly forming sentences in it I can.
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
I wonder if the resultant translation engine could be considered a derivative work of the texts that populated it. This system is standing on the shoulders of all the translation efforts that went in to it. I think it's a great idea, but in the current IP climate, could well be shot down in flames. How much dual-language text is available in the PD or on open content licence?
You know, that actually does sound like something that would be a Russian aphorism...
The use of three languages is critical. Grammar isn't consistant, and words have multiple meanings. By using two known languages, you can eliminate many of the errors thus introduced, because the chances of some error fitting both known languages in the same way is much smaller.
If you double the number of known languages, you more than quarter the number of errors, because although errors can occur in either or both, they're unlikely to be the same error. Once more information exists, you can re-scan the same text and fill in the blanks.
Me, personally - I'd require four languages, three of which were known. The number of texts required would be considerably smaller and the number of residual errors would be practically non-existant.
They chose two languages for the obvious reason: It's simple. It's easy to find a student who knows two languages. At least, easier than finding one who knows four.
However, the price of simplicity is bad science. The volume of information they require makes their system little better than an infinite number of very smart monkeys with text editors and a grep function. That they're being paid signficant money on such stuff is a joke.
If they offered me the same money (and one of those Linux NetworX clusters) I could have a superior system in a month, although (as stated above) it would require more than one known language.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Translating from well-known languages such as Hindi or Arabic are all well and good, but they're already pretty easy to translate (the rules are well-known, translations are easy to check). This may still be a good way to do such translations faster, but it won't help you with: new or obscure languages (not enough translated data to feed into it), quality translation (no mention of the results, so I assume no one's going to be relying on this for publishing or journalism), or very pragmatically dissimilar languages (the rules of conversation rather than grammar). It's a good use of number-crunching, but would you want, say, your wedding vows, free lance article, or software specs done this way?
Hey, i never said it would make a good translation. Just that it could be used.
***WHAT THE FUCK ARE YOU THINKING?***
:)
Look, seriously, even if everyone did speak English, there are still tonnes of literary works in other languages - the original texts of the Ancient Greek classics, for example. To read in the original language is often a much more rewarding experience. Besiders, relying on past translations of non-english material can lead to errors. And consider how many different English translations of the Bible there are.
Almost everyone can speak, read and write at least tolerable english
Almost everyone can communicate using gestures, facial expressions and grunts, but is that any reason to use that as our primary communication method? I mean, to really stretch a metaphor from human languages to programming languages, we can write any computer program "tolerably" in assembler (it's Turing-complete), but that doesn't mean it's the best way to do it. If I can only speak one language "tolerably", but another exceptionally well, which one is better for conveying my ideas?
most young people can have full fledged discussions in it
I don't think we can rely on "d00d, u r so l33t" to teach people true literacy. Young people are increasingly using SMS and online chat and are actually losing their ability to correctly spell words or write grammatically correct sentences. The number of young adults I see who cannot distinguish correctly between there, their and they're is ABSOLUTELY TERRIBLE. Literacy is a major problem in English-speaking nations.
Just look at Slashdot, I'm quite sure I'm not the only one who doesn't have english as primary language
that doesn't mean you can use it well. Take a good look at slashdot - many, many people mangle the English language. The American people are probably the biggest infringers here...
It's not that farfetched idea that in the (near) future everyone uses or at least knows english well enough to make translations meaningless
Human languages don't map to each other 1:1. Some languages have words that basically cannot be translated without a serious loss of accuracy. (I guess you could ssay that no human language is Turing-Complete, in that it can't totally express every conceivable human thought). Having everything translated to english is NOT a solution. Brevity, language tricks (such as puns, rhyming, etc) cannot always be substituted across languages.
If it wasn't 2:15am in Melbourne right now, I'd try to order my thoughts and express them more clearly, but after 4 hours of Java debugging I'm off to get some sleep before uni tomorrow. Goodnight.
The PowerPC includes for this purpose two instructions called SYNC and EIEIO.
BTW, every "% of humanity" statistic has to consider that most humans are Chinese.
I've been thinking about this a lot lately. As in, "Should I be learning Mandarin Chinese?" China is rapidly becoming a high-tech nation. There are a lot more of them than there are of anyone else. Frankly, I think the only thing that keeps China from being the single largest world influence is the fact is its government, and that can't last forever. Sooner or later (maybe after a bloody revolution) China is going to become THE major world power, and the US is going to take a role like that of Great Britian's (no offense to any Brits reading this) today.
End of lesson. You may press the button.
The big problem I see with this scheme is how do you collect the Gigs of data (ie content) without wholesale copyright violation or licensing (big bucks). Sure you can get lots of content whose copyright ran out from the Guttenburg project. But that's gonna be +70 year stuff.
Add the fact that the Mickey Mouse Copyright Extension act and related legislation threaten to extend copyright terms for infinity minus a day and you're never gonna have much content available that reflects CURRENT usage of the languages you're trying to translate.
My impression is that you need the matched texts to *train* the translator. After that, you give it one side and it tries to build the translation, based on what it learned in training (sounds like neural networks and all that to me).
Even existing translation programs could benefit from a ranking system. Wouldn't it be helpful if you could tell just how confident the translator is about a certain phrase or word? That way, you could rephrase your sentence before you foolishly ask someone to "taste" you....
You need an already translated text for his system to work.
Well, you need a pool of already matched texts. Once you have this for a given language pair, you can immediately start translating (presumably). So, it is "universal" in that, once the system is primed, it can immediately begin translating (ie, you don't have to build grammar rules, dictionaries, etc, etc).
True but you could argue that even the english and americans can't speak it well.
Shadus
It's not quite done yet, but the system does show promise. Dictionaries have already been created in Spanish, English, German, Japanese, Italian, French and several other languages.
20 mil and I will! Learn Esperanto with 20M others.
I understand that this is a cool idea for building automatic translators, but is it practical? Basically what they are doing is taking a well-researched domain of languages and trying to make something new and cool in it by completely ignoring the domain knowledge. My intuition tells me that "always use as much domain knowledge as posssible" is an engineering axiom.
I passed the Turing test.
but i think the training only applies to a system for translating between the languages of the datasets used.
so "training" it using parallel texts of japanese and english would produce routines for translating between japanese and english, but not french and english.
As press releases tend to do, this leaves much to be desired for folks who are familiar with the discipline. As I read it, it seems to imply that the main driver is phrase-matching. What does it do with phrases it hasn't seen before? The problem is solved by throwing lots of data at it -- how much data is needed for a reasonable system? How well does it generalize to text outside the domains of the training data?
Incidentally, had my brother been a girl, he was in serious danger of being named Rosetta Stone.
-- Trevor Stone, aka Flwyd
Ceci n'est pas une signature.
I don't think this translation program would be able to deal with his Texan affectations.
I always thought it would be interesting if google applied its page rank algorithm to provide a translation service. Like poll the top 5 translation service sites for a translated sentence and then based on what each of them return, generate a 'average' or best possible result for that sentence.
Right...215.4 million Americans speak only English. 28.1 million Americans speak some form of Spanish (with or without also speaking English). I wouldn't exactly say 13% is "well on the way". English definitely isn't going to be the next Latin. The majority of science, technology, aviation, and computers is English based. It isn't going anywhere soon.
I'm American, yet I speak four languages including my native language.
:-)
What does that make me a quadrilateral?
Just be sure to wear the gold uniform when you beam down -- you know what happens when you wear the red one.
Sure it is... It's not a magic psychic UT like the one on Star Trek that allows you to instantly converse with an unknown species, but it's "universal" in that it has no internal dependencies on specific languages and could be used to go between any two.
Spanish Slashdot: Barrapunto. It's been around for almost as long as Slashdot itself.
In Soviet Russia, Jesus asks: "What Would You Do?"
Wow.
You can simulatneously pimp the Christian bible, lay some smack down on those Christians you don't agree with, and disprove your own point. Impressive.
Seriously, if the many translations have altered the emphasis of the reading, or even the words of the most important speaker, then does it really qualify as parallel data? Heck, I'm only familiar with English-language bibles, and only somewhat at that, and the differences are significant. Imagine the differences (i.e., non-parallel-isms) when Bibles go from Greek to Latin to English to other languages.
(Yes, yes, I'm an American, so I am assuming that everyone works through English. But, I did mod up the joke about people speaking one language as being funny - d'oh I just killed my mod. And, I'm taking the salient part of the LDS comment as true, which would imply some English-language bias at least.)
There are 1.5 Chinese people and 4.5 billion other humans. How do you figure the Chinese outnumber the rest of humanity?
Mac OS X and Windows XP working side by side to fight back the night.
GPG or similar, or using large, one-time pads will always work. Of course, then they just make encryption illegal. What are you trying to hide, eh? Only dishonest people need privacy! Eh!
I want to delete my account but Slashdot doesn't allow it.
When I lived in Europe, a friend and I went to Paris. We're both bi-lingual; myself German, him Spanish, but unfortunately neither of us knew French. We had occasion to ask which train we neededto be on to get somewhere; and asked (in French) if the person we were asking for directions knew Spanish, English or German. We went through a good ten people before we found someone willing to admit that they spoke something other than French.
I'm sure they thought they were being all "Ha-ha, I will not let these Americans get away with not speaking French!" but our interpretation of the situation was "We're americans, we speak two languages, what's wrong with you?"
paintball
20% would be more than any other language, including chinese. So if you were going to pick a language to learn that the most people would understand, you would pick english
For those of you wanting to learn a language that is spoken by approximately 2 million people around the world, start learning Esperanto today!
20 mil and I will! Learn Esperanto with 20M others.
Actually, I think that this may be an interesting way to translate the Bible (assuming you didn't use the Bible itself as a reference...that would skew the translation).
Think about it: every translation of the Bible is always criticized for some reason. If the Bible were translated this way it could be like the Google news of Bible translations: completely independent of human bias and editing.
If it can translate Bob Dylan, then it can do anything.
I don't suffer from insanity, I enjoy every minute of it.
How about piping in various algorirhtms encoded in Pascal and C into the thing and seeing what it does to convert arbitrary sources. Where Can I get the soource? Pawel
Fucking pig ignorant english-centric poster you are.
:)
What's up with this? When did Yoda get such a filthy mouth?
So, can I train this program with a bunch of requirements documents, and a bunch of implementations, and have it learn how to code? :-) If so, I think I am obsolete. *poof*
This automates the task of converting from oe language with a large body of existing translations to another. If you're relying on putting your secrets in French and not having anyone who knows French read it to keep your secrets secret, you're an idiot.
As for used in a court of law, maybe, but who cares? We have an advesarial judicial system - if the translation is wrong, get someone who actually knows the language being converted to/from and refute it.
Speaking in a language that can be translated back to English isn't any more private than speaking in English. There's no loss of anything here.
paintball
Then you "knew him", not "know him".
I'm dead too.
And before that every one that was anyone knew french. Before that latin. This didn't make the need for translation. Basic english may be enouth to find out where the bathroom is or how much a bigmac costs, but many people don't go much ahead of this. You are forgetting that there are countries with many, many people who are not literate in their own language.
And if you are trying to get to know a culture, you MUST know the language. So for works of art, and even for not so artistic content such as movies, a good translation or knowledge of the language of the original country is needed.
[]'s Victor Bogado da Silva Lins
^[:wq
Just think of it.. the most spoken languages in the world are Chinese, then Spanish. English only comes after those two.. and then comes arabic and Bengali.
According to this more people in the world speak languges OTHER than english
*shower*
It makes you a polyglot! :^)
Slashdot's first reaction to VMware
Perhaps you missed this sentence fragment from the original poster:
...I'm quite sure I'm not the only one who doesn't have english as primary language.
It would be hard to classify this person as an American.
http://www.rootstrikers.org/
right... duh.
If you wanted french you'd have to feed it french and engrish
Shadus
BTW, every "% of humanity" statistic has to consider that most humans are Chinese.
If you want to be even remotely close to statistically significant you have to include citizens of India as well most of whom are very different from those of Chinese descent. . In fact most people will probably be an Indian citizen within the next 20 years. However citizens of India are a more heterogeneous population than that of China. Then again, Chinese of the diaspora (eg. in Malaysia, Indonesia, the Philipines, Vancouver etc.) are also a large population but can be very different than mainland Chinese. So I guess in the end every % of humanity statistic that measures some culturally derived phenomenon has to be considered BS.
If you want all those foreign web pages to come back translated into english from 1648.
paintball
I would think the "bloody revolution" would cut down their numbers a bit.
The original poster said:
Now, while this certainly doesn't *preclude* them from being American (there are plenty of Americans whose primary language is not English, after all), I'd suspect they are not. That being said, at least in first world countries the original poster is somehwat correct; most children in first world countries spend some number of years studying English. Whether they are capable of conversing in the language is another story.
However thanks to the huge populations accounted for by the third world, the number of English speakers per capita in the world is fairly small.
Thanks!
This isn't the same Bible that has been exhaustively studied and pored over by scholars over the centuries.
...just my 2 gil.
It is called the Bible. Not only has the translation have been done into many language, but you are dealing with discreet, labelled paragraphs.
Back in 2000, there was a professor at U of MD that was working on using the Bible as a language source.
Fight Spammers!
IBM tried this statistical technique years ago, it's not a new approach. They used the texts of Canadian parliamentary discussion, which is kept in both English and French. See here or just search Google for "IBM tranlslation canadian parliament" or the like.
I is ahead of my time I's been completly ignoring grammer rulz for years now.
***I GOT NUTHIN***
I wonder how this would fare putting two computer languages side by side? I mean... take a few thousand programs, coded using the same algorithms but different computer languages... would his language translation software translate between them? Would it be able to differentiate between languages that manually allocate memory and those that use garbage collection? How about between procedural langauages like C, and more esoteric and oddly structured languages like LISP?
An interesting challenge, eh?
Would there be any benefit to this?
"I will trust Google to 'do no evil' until the founders no longer run it." Hello Alphabet.
In this case I believe the statistical analysis would work. One would just use the 4 other Potter books on the market and their subsequent translations into German. I'm sure JK Rowling writes in a similar style in all the books... so one phrase that means one thing in one book should mean the same in the new one... so for books in a series maybe this should be the first crack? (And then have actual translators correcting what may be wrong?)
Better-informed Americans (a small miniority of the class) would be aware that Spanish is well on the way to becoming the predominant language in the USA.
Why? Because self-identified Hispanics use it a lot?
When non-hispanic subcultures start replacing English with Spanish, THEN it'll have a chance of predominance. As it stands now, Spanish is the largest secondary-language in the nation, and may very well become a national second, but it's not about to replace English.
People have denied the automatability of human skills for the last ten thousand years. "They could not hit a barn at that dista...". Famous last words.
The human brain does not emply magic. It uses strategems, hard-coded guesses, models, and logics, filled with and tuned by accumulated information and knowledge.
All this is automatable. It just takes a _lot_ of investment.
My prediction is that within ten years we will have machine translation that speaks significantly better than average people, although not nearly as well as professional translators.
And it will be as banal as playing chess against a piece of software.
Ceci n'est pas une signature
...and I will make pseudo-insightful comments based on the headline text without reading any of the source articles, until my karma is excellent?
I think what was implied was that if you already had a translation engine trained for English/Japanese, when you are training it for English/French you can use the already existing "metadata" for English/Japanese to make the process quicker (requires smaller datasets to achieve the same precision).
I might be far out here. Excuse my crappy English, btw.
Are you a grammar Nazi? I'm trying to improve my English; please correct my errors!
Now that would be cool.
Seriously though, this leaves only the odd tribal languages of African (and perhaps South American?) tribes that are comprised entirely of clicks and gutteral sounds as not easily comprehended. Could this system's approach finally result in a Babelfish-like universality even for languages such as Chinese and Japanese? The added complexity makes it much more challenging for things like Babelfish, but if this system can do it, it's going to be a landfall discovery.
Anybody have any further research by this guy? I'm interested! Who knows, maybe I could have gotten a better grade in French thanks to this research...
Even though your definitions of "everybody" and "tolerable" differ greatly with mine, at first glance, it appears that you're making a good point. But there are two huge problems with your post.
Firstly, while English is fairly widespread, it is by FAR not any sort of "universal" communications medium. It's my primary language (there's also the Spanish that I rarely get to use, so it languishes pitifully), but I know there are literally BILLIONS of people who don't speak a word of English, and hundreds of millions who have never been--and will never be--exposed to spoken or written English in their lifetimes.
Secondly, even if English were perceived as a solution to inter-cultural communications, the thought abhors me. As ignorant as many Americans are, and as difficult as it would be to get some Americans to consider using ANY other language, English should definitely not be the first choice as a unified "Official Planetary Language."
A worldwide movement to eliminate the use of other languages (or even let them fall to a state where they might meet the same fate as my Spanish skills) would signify an intolerable level of neglect toward so much of the diversity that makes our world interesting and colorful.
Sure, if you speak your native tongue throughout your day and type English when you browse Slashdot, English will continue to be a "secondary language." But you can't deny that if English were the language of choice in every public forum, atrophy would eventually overtake any attempts to retain fluency in one's other languages.
Language barriers can prove frustrating at times, but there are MUCH better ways to improve efficiency. Buy a pocket translator if you're hurting for ideas.
Mom says my
The Rosetta stone itself did not do much in the way of our knowledge of the egyptian language.
What it did do, was provide insight into their method of writing.
It was the latter discovery of the the relation between Coptic and Egyptian that revealed most of the actual language.
(IIRC)
May I point out that the biggest user for this, if it works, are unlikely to be secret services (US or not) or religious folks?
The European Union will have 25 members on the 1 May 2004, if all goes well, and European legislation will then have to be translated into 21 different languages (not 25, some Member States share the same language).
Can you begin to imagine how many battalions of translators we're going to need?
There hasn't been a substantial breakthrough in automatic (i.e. unaided) translation in at least 15 years, and, if Moore's law holds, I'll take this with thanks, as I would be but too happy to throw processing power at it...
I have seen it all: SYSTRAN, originally used to read Russian confidential messages, EUROTRA (sort of son-of-SYSTRAN), METAL (the Siemens system, good if the writer is a cyborg who uses standard building blocks-maybe the oilman in the White House might fancy it...), whatever.
There might be some promise in this approach, as the problems in parsing weird languages (Estonian, for instance) seem at present unsolvable.
This fellow deserves watching, methinks...
Thufir Hawat
Part-time Mentat
You know, I just don't get why christian-bashers tend to be so bleeding ignorant.
I mean, when *I* make fun of someone, I make sure I'm educated enough about them that I don't sound like a total fool.
Maybe the Old King James (or "Authorized Version") was written in the language of the 1600s, but not only is there the New King James, there are plenty of modern-english translations, including those who have completely translated idioms into modern-day English idoms, which makes for much more interesting sit-down leisure-reading than the stuffy old king james version.
Consider yourself educated.
Most of the people that matter speak English :)
The fact of the matter is that the US is the dominant commercial power in the world. They speak a language similiar to English in the UK (but we didn't fight a revolution so that we could ride the lift to go out a light a fag :). Add in Canada, Australia, and the up-and-coming heaveyweight India, and the critical mass gets much bigger. English literacy is very high in Europe. Japan has a failry high level of English literacy, and China does as well - at least amongst academia, r&d, and the international commerce communities.
The fundamental fact is that English is the dominant international language, the language that most people are going to choose to learn in addition to their native language.
Now if the rest of you would just start calling it soccer and drop that silly metric system the world would be even better.
Back to the grand-parent poster's point - since most of the world's technology development and commerce is happening in English already, there is little need for translation. On the other hand, cultural and literary works are still important - but likely cannot be translated mechanically. The middle ground for this technology is basic informational data - news reports, manuals, etc.
Actually, when you consider that most Chinese and a good number of Indianpeople speak English, 50% is fairly accurate.
The current world population (estimated).
Saying Android is a family of phones is akin to saying Linux is a family of PCs.
- Improved Statistical Alignment Models (2000) - Franz Josef Och, Hermann Ney, which investigates and compares several models
- A Syntax-based Statistical Translation Model - Yamada, Knight (2001), which tries to treat sentences structurally instead of just a stream of words
- A Finite-State Approach to Machine Translation - Bangalore, Riccardi (2001), which uses a different way of looking at the problem than usual
Kevin Knight prepared an excellent (if now somewhat outdated) introduction to statistical machine translation that you can see in HTML or RTF (the formatting was corrupted when the RTF was converted to HTML - I recommend the RTF).--
I romp with joy in the bookish dark
That is, I'm afraid, too often the case. The US is well on the way to being a Spanish-speaking country, and if that actually happens, I think it'll be in the best interests of Americans actually being able to understand each other.
Time for inflamatory reasoning. The statistical approach will beat out the grammar and rule based ones, at least for English, is for the simple reason:
English is not a language
Or rather, it resembles one but is more not than is, IMO. It is a large collection of idiomatic expressions that changes quite rapidly (and not only in colloquial forms, just look at what the political-correctness movement has done to phraseology). You know the story... more exceptions than rules, things that are legitimate to say language-wise are considered incorrect anyways, and vice versa, etc. etc.
That's not to say it doesn't have advantages; it's relatively easy to learn the basics of communication since it's weakly conjugated, has genderless articles, fairly simple uncased sentence structure. But, it is monstrous to master and I suspect most native speakers aren't true masters (not to mention the orthographical nightmare; is English the only language with spelling bee contests?)
The reason it's the new lingua franca (or should it be lingua angla now?) is techno-socio-political as is always the case. Stop harping on Americans for being largely mono-lingual. "Why didn't the Romans learn the local languages when they controlled Europe? Because they didn't have to." If every state spoke a different language, which would be more akin to Europe, then there would be need.
Almost everyone can speak, read and write at least tolerable english
/. , because english is the primary language of the site's main audience - stands to reason that english speakers would look at an english language site. On top of that, most people here are geeks, and english is (as far as I know) the human langauge most programming langauges are related to.
;)
:)
That may be the case on
I think it more likely that we'll end up with a lingua franca for the net combining useful bits from any language that has something useful to offer. Look at 'english' as it is now; it's heavily laden with words borrowed from the latin group, nordic languages, even chinese and japanese. And where would we be if you couldn't say karaoke?
Personally, I think it'd be a terrible loss to for the whole species if we all spoke the same language. There are ideas and concepts you just can't express in english that appear naturally in other languages, and I'm sure english has much to offer people who don't speak it as their first language.
Variety in all things...
Warning: May contain nuts
If you want to transmit something secretly then encrypt it! If you idea of secret communication is speaking in another language then you sorely need to learn more about cryptography.
I had this idea a few weeks ago after reading a biography of NEC founder of Koji Kobayashi, whose dying wish to NEC engineers was to have a machine that could instantly interpret English and Japanese speech by the date of his hundredth birthday in 2007. I wrote down my brilliant idea to use only statistical matching and a huge database of texts to make a translation on a napkin from the coffee shop I was in. I spent a few days thinking about it some more, but decided I would put it off until I had a better computer (better than my P266MMX) and could actually program. I guess I can forget about that now though... Sigh. And to think this guy had me beat before I'd even thought of it.
You may have methodological problems with their approach, but it works. Simple approaches are not fundamentally bad, specially simple approaches no one has tried before. From the small amount of information available, it looks like a very promissing path once you have enough storage and processing power. Guess what, we now have both at consumer level prices. So why not try the "dumb" method? Specially if it works better than all other methods available.
Good luck getting a many-gigabyte database of Old Greek and Hebrew texts with translations. And better luck getting such a database that does not include the Bible in its contents.
Finally! A year of moderation! Ready for 2019?
Most Chinese don't speak "Chinese" either. They can read a common language, but there are hundreds of mutually unintelligible dialects in China (at least as distinct as, say, Norwegian and Swedish, and often more so; Croatian and Serbian are just the opposite - mutually intelligible spoken language, different writing systems). The definition of what counts as a language is often politically motivated, so lumping all of China into "Chinese" and assuming all India fits into Hindi makes for some skewed numbers.
With that in mind, English ranks higher in number of native speakers than you might think, and is way, way ahead in number of non-native speakers. This is especially true for written language (where machine translation can somewhat work).
I have no idea why you think Spanish is going to be the predominant language in the USA. My guess is that either you have birth rates and # of language speakers conflated or just feel that even the slightest trend toward bilingualism requires a winner and loser.
... where the more available examples of actual spam and actual non-spam the better the accuracy of the result, and where you basically let the computer work out the probability, rather than feeding it hard and fast rules up front.
Can anyone say if the two procedures are technically related?
Actually, if you look at the notes on the statistics, they are only the first language speakers in each country.
From a Time article last year:
Mandarin may have the largest number of native speakers (about 800 million), but English, with 1.9 billion speakers--including some 350 million native speakers--is far and away the largest global lingua franca. The next largest, Spanish, claims 450 million competent speakers worldwide, while French is spoken by a mere 130 million. The most vital statistic is that some 1.5 billion people around the globe speak English as a second language. "It has become the working language of the global village," says ESU chairman Lord Alan Watson.
For native speakers, Chinese wins. For overall comprehension, English is out in front (which doesn't affect the need for good translation tools).
"You look nice..." --> "Shall I compare thee to a Summer's day..."
-- Sig down
Exactly how is 20% 'most people'? If I have 20 green marbles, 10 black marbles, 5 blue marbles, 15 orange marbles, 9 plaid fuschia marbles, 11 clear marbles, etc., are 'most' of my marbles green? No, I have more green marbles than any other kind.
So that would mean (if 20% is indeed higher than any other language, I have seen no statistics to verify this fact -- last I heard one of the main Chinese dialects was the most widely spoken language in the world, not English, but this was 20 years ago) if you were going to pick a language that was (based on what you have said) the most widely spoken language, then English it would be. But my own understanding is that this is not the case.
My journal has hot
I know this is a dramatic title. But better translation systems will do as much for cross cultural communication as the internet has done for cutting through geography.
I am waiting for the day when I can read Middle Eastern texts without having them selected, censored, and biased by american publishers and editors.
Most of the people that matter speak English
Okay, so 80% of the world population is irrelevant?
My journal has hot
Because of the large volume of required input language into the system, I don't think that this system will be good for translating Bibles into new languages (think Wycliffe USA).
The advantage of this system, as it would pertain towards helping a particular religious community, is to make it easier to translate the large amount of books on religion into other languages more accurately.
The Czech schoolboys who worked so hard to make a Czech translation of the Harry Potter books would no longer have to wait so long for translations. Rather, the book could be fed through the statistical system and in under a day (minus proofreading) there could be a very nice translation into any mainstream language you wanted.
The advantage here would not be in translating Bibles into new languages, but rather translating massive amounts of books on the subject of Christianity to various languages.
Perhaps this technology would even have a use with Amazon's online digital book project in allowing all of the books on their site to be effeciently translated into other languages and marketed digitally. Interesting concept, to have all of the resources of Amazon.com in digital format for any mainstream language. That could do amazing things for the cross-country circle of ideas and thoughts.
Just my 3.14159 cents...
Respectfully, :)
clint
The Rosetta Stone translates YOU!
Human style translation is nice. Machines are starting to be programmed to tackle problems like normal people do, not just like programmers do. About time.
I've had the same thought, but wonder if it should be Cantonese -- the south (near HK) speaks Cantonese and is at the forefront of Chinese high tech and economic development.
We went through a good ten people before we found someone willing to admit that they spoke something other than French.
And what made you think they did actualy speak anything other than French? I'm French and I can say from personal experience that the percentage of adults who can speak anything other than French is quite low. Sure we all learned two foreign languages in school but after a few years most of us don't remember anything... about the only thing I remember from my Spanish class is "me llama guillermo" and I'm not even sure this is correct (it was 19 years ago).
I find it funny (kind of) when an American think of Frenchman as arrogant because they don't speak English... I guess what you're really thinking is : "Ha-ha, I will not let these Frenchman get away with not speaking English in France!".
So my interpretation of the situation is "what's wrong with YOU"
It's the PARISAN French.
;)
French people were very nice/helpful everywhere else I went in France OTHER than Paris. As it turns out, the rest of France doesn't much like Parisans either.
paintball
In short, yes.
Please understand that I am not saying that I don't have compassion for that 80%, or that they don't matter to me on an moral/ethical level.
I don't have an direct interaction with those people, however, and so it does not matter to me whether or not they can speak English or I can speak their native language.
People are needlessly critical of others for not being multi-lingual. But the fact of the matter is that for most Americans multi-lingual doesn't do anything to aid them in their lives. Europe is different, of course, because of the large number of relatively small nations with their own languages. A Dane who works in Sweden and vacations frequently in Germany will derive worthwhile benefit form knowing Danish, Swedish, and German. A New Jersean who works in New York and vacations in Florida derives little benefit from knowing non-English languages.
The fundamental fact is that the international communities with whom I am likely to interact probably speak English already. It may not be fair that they learn 2+ languages and I only need to learn one, but life isn't fair.
I don't expect anyone in a non-English speaking country to speak english. (Hell, sometimes, I find that expecting even native-born Americans to speak english is a bit much.)
The amusing part is that most of the people we talked to almost certainly knew one of the languages that we did, but preferred to cop the Parisan "I'm better than you because I know more than one language!" attitude - not realizing that they were essentially pretending to be stupid, not conveying that they were 'enlightened' like they were trying to.
paintball
They've had the same technology at CMU's LTI for years now, called EBMT. This officially stands for Example-Based Machine Translation, but those of us who worked with it called it Extremely Bad Machine Translation because it took millions of example sentences before it started to not suck, and even then it required manual tweaking and the addition of primitive grammar rules.
So yeah, this method learns fast, but it generally learns to a useless level for anything other than a rough assessment of some of the phrases that were in the original text.
-- Fratz, human
I know enough (or at least I did back then) of Germanic/Romance language roots for transportaion-related stuff to read train schedules in pretty much any (Western) European country. There was just some particular nuance (coupled with the fact we were in a bit of a rush) that made asking necessary.
Another trick is to make use of the automated ticket purchasing kiosks. They generally let you make your purchase in a few different languages. Problem there is they can't get you the full range of iteneraries (wierd connections, stops, layovers, etc) you can get at an agent; but the agent may not know a language you can speak...
So we would just buy tickets to somewhere in a language we knew, cancel the order, do it again in the local language, write down the words that corresponded to the words in the previous language, and then just present that order on paper to the agent. Worked pretty well.
paintball
The idea seems very similar to Eli Abir's, now commercialized at the company of
Fluent Machines.
jpenguin AT the google email service
a unilingual anglophone. (say that ten time fast)
A compiler (or interpeter, for those of you into that sort of thing) takes programs written in your preferred language and translates them to machine code.
The benefit to using compilers allows you to see what language/compiler produces the most efficient code. Ideally all compilers should be able to produce the most efficient code, but they'll each have their own strengths depending on what they're designed to do.
None of the languages you give were designed for ease of learning. Nor are they free of a whole load of cultural baggage.(1) The concept behind Esperanto is to provide a neutral, quick learning experience for newcomers.
-------
(1)Now, Esperanto does have a decided tilt towards European languages for its base, so I'm sure there's room for improvement for inclusiveness. But at least it's a higher level of cultural bias.
Due to unprecedented developments in technology, A call was sent out today from DARPA and MIT jointly to all aspiring sci-fi writers and directors to get off their butts. Having now created preliminary versions of just about every StarTrek, Asimov, Clarke, etc device possible, DARPA and MIT both are running out of cool ideas and will need to revert to evil geshhhtaaaapo technologies if we can't find something cool to work on.
Joe Schmo said, "Keep a scientist off the streets, write a story. WIthout cool sci-fi to keep them up at night they'll be building super-world-destructo bombs, and other "evil genius" devices..."
meh
You need to get my church to invest in the latest versions.
Although personally (not like I've done an exhaustive comparison) I've found that the more modern translations lose a lot from the King James version. Seems like the goal with the King James version was accurate translation, while most of the modern translations are geared towards reaching the masses.
Regardless, I wasn't Christian bashing. It was meant to be half-funny in the "Oh look, we got the wrong english" sense, and half-serious, in the "using a particular text where the 'timestamp' for the known languages used may be different by hundreds of years, and where one 'known' language's translation may be based on an older translation from another 'known', but not source, language, might not work as well as you'd hope." sense.
paintball
That's quite insulting to the millions of Americans to whom English is a second language. The 1950s view of American monoculture is dating fast, at least down the coasts.
For starters, we specifically target young people when asking questions where a non-native language will be required. 3-4 of the people were employees, indicating at least a passing knowlege of "What track is this train on?" in a few European languages might be a job-relevant talent. Additionally, the sneer. Attitude is attitude regardless of what country you're in.
We don't expect people to know foreign languages. We *DO* find it amusing when people who are razzing *US* for not knowing THEIR language do not know any foreign languages.
paintball
Given the impressive progress made by Bayesian algorithms in spam detection, I wouldn't be surprised to see impressive results from this method either.
So bravo for Franz Och! He's taken what appeared to be an intractible problem requiring magic AI to solve, and perhaps found a way to solve it effectively using the stupid brute force methods computers are so good at.
I don't care if it's 90,000 hectares. That lake was not my doing.
This program can be made better if it first builds the grammer and dictionary as a foundation. Then, it gets fed with the parallel text data to build the statistical references between the two languages.
It is the combination of definition and context that makes a translation more accurate. In fact, that's how humans learn. We first learn to reference words/sentences to what we see (like parallel data). Then, we also learn to understand grammer and word definitions. We wouldn't learn by just one way without the other.
On another note, I think this will be great step in achieving speech AI. Just as this program translate by making statistical parallels between language, it can go further and make statistical parallels responses to a sentence. Reason why I think this program is better is because it goes beyond plain definition, the program has to "understand" the input text by searching thru its database, its "knowledge". It's still far fetch, but i think this is one step further towards AI.
Ouch. Don't need a translator for that.
1: Create a set of "Rosetta Stone" data by taking thousands of recorded phone calls to customer service/operators, etc.
2: For each call, track what the customer service rep/operator typed into their computer terminal.
The result would be natural language voice-recognition that would probably achieve a high degree of accuracy because it would be limited in scope (e.g. asking for a credit line increase, reporting a lost card, checking your balance, etc.) and be based on real queries from real customers.
Since the biggest majority of calls are for very simple problems (I forgot my password is the most common tech support call we get) this should be pretty useful.. you could probably automate "Level 1 Tech support"!
The facts have a liberal bias. --The Daily Show
Ah, so the DoD must be secretly staffed by slash/shonen ai-loving women. I get it now.
--grendel drago
Laws do not persuade just because they threaten. --Seneca
Some people have decided to use Slovio (a language) on their porn sites. Their idea is based on the fact there are 400 million potential customers of which only a very small part is speaking English.
This has already been done some years ago in Canada, where the translation system was fed the complete text of parliamentary debates for umpteen years (required by law to be translated by humans into French, if originally in English, and vice versa). I don't know how it fares when presented with a sample of parliament-speak (I concede, this is not a fair approximation of human language), but it fails miserably on a simple rhyme. Read your Hofstadter, guys.
I can assure you, the best way to get rid of dragons is to have one of your own.
Ever hear of that wonderful American invention called an airplane? How about that other American invention called a telephone? What makes you think you need to drive anywhere to have use for a foreign language?
Do I really need to learn another language?
You don't need to learn anything. It's not as if billions of people are finally getting the education you take for granted in the First World, and are just dying to take your job. Nothing to worry about.
With Mandarin you can get by almost everywhere
in China that would have tech higher than an oxcart.
The schooling in China is done in Mandarin. Public
schooling is mandatory. As a result essentially everyone
in the country below the age of 50 is fluent in Mandarin.
Street life in HK, GZ, SZ, is in Cantonese (less in SZ, since
it is so heavy with migrants), but business, science, tech,
will use Mandarin primarily.
-I like my women like I like my tea: green-
"I'd hazard a guess that this system will also have trouble with a high-context language such as Japanese."
:-)
I think that's exactly what his system is about, it also analyzes context, so it works similiar to the human thinking. It takes the surrounding words into consideration and checks given translation.
E.g., wrong translations for the German word "Bank" which can mean both "bench" or "bank" are less likely, as a context like "sit, park..." would favor the translation "bench" while a context with "money, stock..." would lead the system to assume the meaning "bank" is more likely.
This way, it should be far superior to word-to-word based translation systems.
I hope this is true, but that's how he explained it to me about 7 years ago when he started the project
I don't need a signature.
goodness, i certainly didn't think that when i first logged into /. today that i would be seeing old english...
it's astonishing just how germanic it looks to me.
ed
Here are some things that language allows, but reality doesn't:
IANAL (linguits), but you also run into peculiarties of language where on language lacks a concept. English does not impart gender in objects. Ancient Chinese is largely written in the present tense. Some african languages have no concept of "Should". In order to translate information you need to add, remove, or complete rephrase certain ideas.
Finally, you run into the problem of ambiguous concepts. There is no german word for "luck" or "happy". They are combined in one word "glucklich". The chinese use one character to represent both danger and oppertunity. To know which you are speaking of requires context, and once you start adding state to a statistical model it starts to become a differential equation.
A far better approach would be a digital "Esperanto". Linguist would design a universal language, and then design a filter to translate each language (and all its quirks) into the universal. Each language would also need a filter to translate FROM the universal. Even then, you still would have stuff that just plain doesn't translate.
For giggles, try picking up a copy of Sun Tsu's Art of War, or Lao Tsu's Tao Te Ching. Better yet, pick up 2 different translations. In order to make any sense out of it, you have to constantly read between the lines.
Computer are notoriously BAD at reading between the lines.
"Learning is not compulsory... neither is survival."
--Dr.W.Edwards Deming
um, they have 'em. they're called parallel bibles, are available on CD-ROM and the most complete ones include both the original aramaic & latin texts along with hebrew and other translations to boot, like this one: http://www.powerbible.com/
ed
The `spirit is willing' story is amusing, and it really is a pity that it is not true. However, like most MT `howlers' it is a fabrication. In fact, for the most part, they were in circulation long before any MT system could have produced them (variants of the `spirit is willing' example can be found in the American press as early as 1956, but sadly, there does not seem to have been an MT system in America which could translate from English into Russian until much more recently --- for sound strategic reasons, work in the USA had concentrated on the translation of Russian into English, not the other way round). Of course, there are real MT howlers. Two of the nicest are the translation of French avocat (`advocate', `lawyer' or `barrister') as avocado, and the translation of Les soldats sont dans le café as The soldiers are in the coffee. However, they are not as easy to find as the reader might think, and they certainly do not show that MT is useless.
BTW, since this book is no longer available in the stores, the whole contents is placed online, though the server appears down right now. I recommend reading this book to anyone who is interested into the subject of MT. It really is a nice introduction into the subject.
You mean that the Hebrew text wasn't the original?? Oh my GOD!!!!!!!!!!
- Marco
wish i could edit...
i didn't wanna lump it in w/ the other stuff in the list: it's sorta in a different class by itself, IMHO.
ed
Neither the approach nor the results are particularly unique: statistical models are all the rage in natural language processing--and for good reasons. There are probably a few dozen research groups working on these kinds of systems.
But such systems also have well-understood limitations. Translation often does require a deeper understanding of the subject matter, and you simply cannot get that from aligning two large corpora.
Arabic, Chinese, and Hindi? It is pretty easy to imagine why DARPA would be interested in the first two with the War on Terrorism and China being a pseudo-enemy. But Hindi? Conspiracy theorists please comment...
You don't get the situation. French feel enormously threatened by the Americans, culturally speaking. Maybe the feeling is even founded on something...
If you approach them with English, or even saying "Do you speak English?" in broken French, they would rather watch you being eaten live by sharks then understand your cries for "Help!".
What you do is, ask the French person what-ever you need to know in your best French. Probably he will not even understand what the hell you're trying to say. Speak to him (in French) again. And again. Bring out your phrase book and shuffle the pages, speaking uselessly at the french guy.
Maybe he WILL understand parts of it. This will just cause him to unleash torrents of speedy French in your direction, which you can't understand. Look stupid. Stupidity is key.
Soon, it will be obvious that you don't know shit about the french language. NOW, and ONLY NOW, ask in your very brokenest French (no, NOT English, you will ruin everything you just accomplished!) if maybe he knows just a little English.
Of course he does. After all, all the kids study English in school, and have so for many years. This way he can feel superior about helping the daft tourist idiot.
NB. I am not american, I'm swedish. I had the exact same problem in Paris until I developed this method.
I choose to remain celibate, like my father and his father before him.
can't we all just speak Esperanto?
(i suppose i should've written this message in esperanto to illustrate my point, oh well)
THERE IS NO DATA. THERE IS O
Yeah I know there was an Esperanto movie. Exactly my point of free from culture.
I just had to. Besides, I think it's proving a point, or something.
--
Romancing of the Rosetta stone
' you give me sufficient parallel data, and you can have translation a system in the hours '
University southern California of the computer scientist Franz Josef, which Och of most famous against-resounded, praises itself in the history of the technology, after its software counted the Arab strongly under 23 and Chinese English translatio systems, commercially and experimentally, examined inside in recently concluded Ministry of Trade of attempts.
"you indicate a place to me to the location, and I shift the world,", after to to order a mathematical explanation for the lever said the large Greek scientist Archimedes place.
"you give me sufficient parallel data, and you can have translation a system for all possible two languages in an affair of hours,", said Dr. Och, a computer scientist in the USC school of the institute for information science of the technology.
Och spoke after the benchmark tests 2003 for the machine translation, which was accomplished in the May and June of this yearly by the National Institute of Standards and Technology United States of the trade department.
Translations Ochs examined well into the 2003 head ton head tests against 7 Arab systems (5 research and 2 commercial away dregal products) and 14 Chinese systems (9 research and 5 from stock). In preceding 2002 evaluations had examined it similarly superior.
The researcher discussed his methods held at a NIST Postmortemseminar over the Benchmarking July 22-23 of John Hopkins at the university in Baltimore, Maryland.
Och is an outstanding exponent of a newer method of using the computers to touch in order to translate a language into other one, which became more successful in the last years, while the ability of the computers grew, large bodies of the information, and the volume of the text and the brought together translations in the digital form has, on (for example) multilingual newspaper or government net places of assembly explodes.
Method Ochs uses brought together bilingual texts, the computer-coded equivalents of the famous Rosetta descriptions of stone. Or rather gigabytes and gigabyte Rosetta of stones.
"our approximation uses statistic models, in order to find the most probable translation for a given entrance," Och avowedly
"it is rather different to the older, symbolic approximations for the machine translation, which in most existing the commercial systems is used, which try, to code the grammar and the encyclopedia of a foreign language in a computer program the grammatical structure of the strange text analyzed, and produced then English, which on hard guidelines," it is based, continued.
"employs, explaining from the computer, how one, we left it it out explains translated. First we draw the system it with a parallel korpus i.e. an accumulation of texts in the foreign language and their translations into English.
"the computer uses these information, in order to co-ordinate the parameters of a statistic model translation of the process. During the translation of the new text, the system tries to find English sentence which is the most probable translation strange entrance of the sentence, be based in these statistic models."
This method ignores or rolls over rather, finds express grammatical guidelines and even traditional dictionary lists of the vocabulary in favor of leaving the computer matchup samples between given Chinese or Arab (or any another language) texts and English translations.
Such abilities grew, while computers improved, by making possible for them, from using the individual words as the fundamental unit on using the groups of words to move -- cliches.
Versions of the different human translators of the same text change frequently considerably. Another key improvement was the use of repeated English human translations to permit the computer too its transmission by an ana
"Mind, as manifested by the capacity to make choices, is to some extent present in every electron." -Freeman Dyson
A bit? You got that right. Let's say that, in the course of the hypothetical Chinese revolution, 50,000,000 people are killed. This is roughly the number of people who died in World War 2. That's a lot of people, well over a hundred times the entire population of the USA.
This brings the population of China down from about 1.5 billion to about... 1.45 billion.
The phrase "drop in the bucket" springs to mind.
End of lesson. You may press the button.
3-4 of the people were employees
I think it's a mistake to judge all Frenchmen based on your experience with government's employees. Particularly the one from SNCF! Everyone will tell you they are the worst!
Ok... Jokes aside, most of them are good people but we always remember the arrogant bastard. Also the truth is a lot of people blame them for whatever reason and they tend to be very defensive.
Additionally, the sneer. Attitude is attitude regardless of what country you're in.
Attitude is closely related to culture, so you must never base your interpretation of an attitude on your culture. For example if you met me one day, you would probably think I'm a cold person : I won't greet you with a big smile and certainly won't invite you for dinner with a tap in the back. But it doesn't mean I despise you! It only means I don't want to impose myself. For me, the guy who greets me with a great smile, shake my hand and invite me to dinner the first time we meet is a self-centered asshole.
We *DO* find it amusing when people who are razzing *US* for not knowing THEIR language do not know any foreign languages.
Well, I'm sure you know that most people in France never use soap... Same kind of thing. It's always fun to bitch against someone different so we can feel good about ourselves. Don't take this to seriously.
Interesting method.
It seems to me this is more similar to natural learning of a language (usually at a young age) by exposure and immersion, as opposed to scholar learning of a language in classrooms, etcetera.
It shouldn't be surprising that in humans, the first method also works best at acquiring fluency in multiple languages. As a matter of fact, it's the only method through which we come to understand our FIRST language, which is in almost every case the one we command the best.
I think most people get, by consuming huge amounts of information, a feeling of "what sounds right" and "what sounds wrong" that is more effective for them at predicting the unwritten rules and exceptions, both in translations and in original sentence-creation, than memorizing a set of grammar rules which, in the end, are just codifications of the current state of the language.
I don't think the success of the approach means the symbolic methods are pointless for this endeavor, any more than the formal study of languages and their grammars is for human translators.
Professional writers and translators do study such rules to dramatically improve their command of the different languages, and do get much better results.
But it seems to me they are more successful going from "statistical matching with massive real-use data" to "optimized grammar rules matching the data" than going backward, from "scholastic grammar rules" to "consumption of massive data to acquire exceptions, and correct and complement the rules".
What would be interesting, I think, is if one can study the state of the system after it's performing well and extract/deduct grammar rules, algorithmically.
It would be interesting to see the results of a program doing that, collecting (and correcting) the grammar using the data, and using the grammar rules when no match in the dictionaries is found to, say, apply a greater weight to the gramatically-correct choice among the alternatives.
If the results were good with this approach, one could consider decreasing the size of the database as the grammar gains stability. Use that memory for other processes, other languages, or new sample data that could not be examined before.
Freedom is the freedom to say 2+2=4, everything else follows...
This software would operate on a deeper level than it would if it operated with the words and symbols themselves. It would utilize a map of the deep structures of language, instead of a map of the less-meaningful words and grammars.
Actually, as a result, it operates on a shallower level. In fact, it's almost like you wrote this comment for an article in a parallel universe where statistical translation was the norm, and somebody was just now proposing symbolic translation, so much so that it's almost spooky.
This translation technique is so shallow it doesn't even particularly care what languages it works with. In a way, it can't really be said to be "translating" in the traditional sense; it's just correlating phrases with no clue what they are.
Traditional symbolic translation is better described by what you said:
Therefore, instead of playing with messy grammars and sentence structures, we can simply have a catalogue of thoughts as represented by words, and correlate that catalogue with a different set of words to facilitate translation.
Word(/phrase) -> symbol -> word(/phrase) is traditional tranlation. This is word -> word translation.
It's working better because we've had little or no success creating the middle part of the symbolic translation; matching the symbology used in our head has proven impossible to date. This works better by skipping that step, which introduces horrible distortions by forcing the words to fit into an incredibly poor symbology (compared to what we're actually using).
However, in theory, traditional translation should still have a brighter future; this is a hack around our ignorance, perhaps even a good one, but eventually we will want to extract the symbols.
(Incidentally, it's also why this same technique can't be used to match words -> symbols; we don't know how to represent the symbols yet! This kind of technique could eventually potentially be hybridized with something else to attack that problem, but simple, direct application can't result in the complicated relationships between symbols that exist, and we'd want a computer to "understand" those relations before we'd say it was truly translating or understanding English.)
Anyways, just flip your comments around 180 degrees and you're pretty close.
My point was that for this program to work you need a whole bunch of text translated between the two languages. This bunch of text must not already contain the new stuff you want translated. So try to find a lot of Aramaic and Old Greek translated to something else, without using anything from the Bible. The article says you need gigabytes.
Finally! A year of moderation! Ready for 2019?
England is not the language
or something, similar one but it compared to it are, the IMO. Is idiomaticas expression was not more great collects its absolutely rapid change (and not only by familiar form, right glance does movement this politics corrects fraseologia). You compared to know to historical... are more exceptions is the rule which legitimately thinks, the matter language wise person is considered incorrectly in any event, with vice versa, and so on.
One does not have not to think its advantage; Its relatively easy academic society correspondence foundation because of it by conjugate weak, to be had article genderless, quite simple uncased official lecture structure. But, he are strangely dominate and I suspect the most localities are partial I are not the real master (do not mention this orthographical nightmare; The only language and the spelling bee's competitor is English)
This reason is frank new lingua (or it must now be lingua angla) he always is the techno partner politics likely case. Stops harping uses the American is to big scale scope monolingue. " why does Romans do not have the academic society place language when they controlled Europe? Because they do not have. " if each condition speaks one different language, that with Europe related, there then is necessary.
Flourescent (adj): smelling like ground wheat.
Correct spelling in Slashdot? You gotta be kidding, they would revoke by club membership.
(I hope it is obvious I just thought it was too small a comment for me to need preview - the most common fatal error. But I haven't started anything with "end". What do you think the two "..." mean?)
heh, the french have the same joke, starring themselves as the latter party.
People replying to my sig annoy me. That's why I change it all the time.
I wanted to use Babelfish to do a English-Russian-English translation on that phrase. Imagine my shock when I discover that while there's a "Russian to English" option, there's no "English to Russian" option.
;
"That can't be right," I thought. So, I took the code from the "Add Babel Fish Translation to your site" link, created a web page, pasted it, and added the following option to the list:
<option value=en_ru>English to Russian
Wonder of wonders, it worked! I have no idea why Babelfish isn't displaying the E-to-R option, but this is a functional workaround.
In case you were curious, the results were disappointingly mundane. "Spirit is willingly ready but flesh it is weak."
Here's the complete HTML document.
<html>
<head></head>
<body>
<FORM ACTION=http://jump.altavista.com/searchbox4.go name=mfrm>
<input type=hidden name=doit value=done>
<table width=200 border=0 cellspacing=0 cellpadding=6 bgcolor=#93b2dd><tr>
<th colspan=2 bgcolor=#FFFFFF><a href=http://www.altavista.com>
<img src=http://a12.g.akamai.net/7/12/282/13/av.com/sta tic/i/af/box_logo.gif border=0 width=118 height=45></a><br>
</th></tr>
<tr><td colspan=2><img src=http://a12.g.akamai.net/7/12/282/13/av.com/sta tic/i/bf/Bfishheading.gif width=192 height=20><br><font size=2 face=arial,helvetica,sans-serif color=#FFFFFF>
<small>Type or Paste text or Web address<br> (beginning with http://) here:<br> </small>
<textarea cols=20 rows=2 name=urltext></textarea>
</td></tr>
<tr><td colspan=2><font face=verdana,arial,helvetica,sans-serif size=2 color=#FFFFFF><small>Translate from:<br></small></font>
<select name=lp>
<option value=en_zh>English to Chinese
<option value=en_fr>English to French
<option value=en_de>English to German
<option value=en_it>English to Italian
<option value=en_ja>English to Japanese
<option value=en_ko>English to Korean
<option value=en_pt>English to Portuguese
<option value=en_es>English to Spanish
<option value=en_ru>English to Russian
<option value=zh_en>Chinese to English
<option value=fr_en>French to English
<option value=fr_de>French to German
<option value=de_en>German to English
<option value=de_fr>German to French
<option value=it_en>Italian to English
<option value=ja_en>Japanese to English
<option value=ko_en>Korean to English
<option value=pt_en>Portuguese to English
<option value=ru_en>Russian to English
<option value=es_en>Spanish to English
</select></font></td></tr><tr>
<td><input type=submit value=Translate style=font-family:sans-serif;font-weight:bold;colo r:#FFF;background-color:#990000;cursor:hand;margin -bottom:-1px;width:85px;></td>
<td><font face=arial,helvetica,sans-serif size=2 color=#FFFFFF><small>Powered by Systran</small></font></td>
</tr></form></table>
</script>
</body>
</html>
'Almost everyone'? What *are* you talking about? You must be an American. From a recent online Harris poll, most Americans think at least half the world speaks English. This is just plain wrong. The truth of the matter is that it's more like 20%. That's it. Most people on the NET might speak English, but most people in the world? Hardly.
By that same token, every country that has an international airport has English speakers in the tower... The same is not true of Chinese.
"Your superior intellect is no match for our puny weapons!"
What? The USA has around 290 million people. How is fifty million over a hundred times the entire population of the USA? Unless I'm misunderstanding something...
"Nothing shocks me. I'm a scientist." -Indiana Jones
Very well-said, raehl. Based on my own trip to France in February 2002, I have a some things to add...
1. Do the French hate Americans?I was told in advance by several people that the French look down upon everyone else--especially Americans--with a certain air of distaste. For the most part, I didn't find this to be the case. As long as you remember you're not at home (America, in my case)--that you're a guest--and to behave accordingly, there's no problem. I saw a few people staring as if they didn't want us there, but not many.
I'm a pretty gregarious guy, easy going, laid back. For example, saying "Bonjour" with a friendly smile before asking for help came naturally to me. I would be surprised if my accent fooled anyone
but I think they appreciated the effort. No doubt they could tell immediately that my "Parlez-vous Englais" was stilted by a from-across-the-Atlantic accent, but they also realized I respected the fact that I was in THEIR country. I may have been "just lucky," but nearly everyone I encountered in Paris was "just fine," attitude-wise. We enjoyed spending 30-45 minutes just chatting with the guy who managed our hotel the last three nights.
2. It's a major city. Go to any large city in the world, and you'll find people with ugly personalities. New York, Los Angeles, Tokyo, etc. Their local economy is largely tourism-based. I'm sure some people don't mind that fact, but others resent it, yet they continue to live there. And some people are just plain assholes. Any large city is bound to have a higher concentration of them. Also, to paraphrase a point someone else made today, you remember the one jerk out of ten nice people.
3. Are the French arrogant? I wonder if a large part of this perceived animosity stems from a phenomenon becoming increasingly prevalent in the United States; many of the people I meet here at home have a mentality that drips with egotism. They're not bashful at all when expressing the idea that they're not treated well enough by everyone else, even when nationality is a non-issue. I can only imagine trying to deal with this type of person and trying to remain civil.
I didn't get this impression from reading posts by raehl, by the way. I'm betting that he just ran into the wrong people. And raehl makes a valid point about expecting fluent English from the average American.
Going to the mall with a girl I used to date: When she was behind the wheel, she would impatiently complain about people walking too slowly in front of her as she circled the parking lot to find the absolute closest spot, and if she could get by without waiting, she would. Yet she was intolerant of motorists who didn't slow down to let her walk in front of their vehicles. Clearly her mode of transportation wasn't all that was pedestrian. More and more Americans have displayed this sort of behavior. I don't know if it's worldwide, but I have a feeling it's not. The only reason this feeling bothers me is that it frightens me to be part of THE country known for one-sidedness such as this.
Incidentally, the worst of the handful of bad experiences occurred when a woman approached me outside the train station a few blocks from Notre Dame. She asked me a question that sounded like a French request for directions. I asked her "Parlez-vous Englais?" (am I spelling that correctly?), After a moment of shock, she looked at me as if I had just attempted to rob her, and then she stormed away.
One of the nicest experiences started the same way: My friend and I were asked for subway assistance on Champs Elysees by two attractive young French women visiting Paris for the first time (sorry, no "dear Penthouse" story here). Like the other woman, they clearly thought I was French; maybe it's because I don't own any Hawai
Mom says my
Another one that gets mentioned is "Out of sight, out of mind" which gets reverse-translated into "Invisible idiot".
Ceterum censeo subscriptionem esse delendam.
The whole world is literally at your finger tips. Here are a few examples.
20 mil and I will! Learn Esperanto with 20M others.
So i won't try. (-;
20 mil and I will! Learn Esperanto with 20M others.
And you think it's bad in English? Try it in German or French where the governments are trying to control language evolution with legislation, instead of just textbooks.
The Voynich Manuscript!
And where are you going to find gigabytes of parallel Klingon-English texts?
No seriously, this is the fallacy behind any statistical approach to automated translation.The news release gives the telling comment:
This paragraph just doesn't make any sense to me. Either it's badly explained, or the entire approach is flawed:
Statistical methods just cannot deal with the subtlety of meaning to be found in natural language texts. It's a little like believing that you can always win at chess if you can just look ahead far enough. I believe that this approach is inherently limited and any apparent success is illusory. This news release hasn't changed my opinion.
Sorry to be a party-pooper, but that's how I feel.
Bizarrely enough, I just heard that joke being given by a man on the street about 2 hours ago on the Paramount Comedy Channel. Freaky.
Where's this great title we were supposed to find?
I had the same problem - having lived in Europe for a year, I was not wearing the typical brightly-colored or college/professional sports team-themed American tourist clothes, and was frequently asked directions etc. by the locals, or other tourists assuming I was a local. Always a blast when someone from the British Empire approaches you and asks in the local language if you speak english.
;)
And for the record, once we were outside of Paris, virtually everyone we met in France was great.
But next time, you should put a little more effort into it when two attractive women ask you for directions.
paintball
its not a majority, its a plurality, more than any other faction. Any percent can be a plurality as long as its more than the other factions. I never said that most people learn english, I merely said that english is the logical choice to learn if you want to be able to communicate with the most people.
most ( P ) Pronunciation Key (mst)
adj. Superlative of many., much.
1.
1. Greatest in number: won the most votes.
2. Greatest in amount, extent, or degree: has the most compassion.
2. In the greatest number of instances: Most fish have fins.
n.
1. The greatest amount or degree: She has the most to gain.
2. Slang. The greatest, best, or most exciting. Used with the: That party was the most!
pron.
(used with a sing. or pl. verb) The greatest part or number: Most of the town was destroyed. Most of the books were missing.
adv. Superlative of much.
1. In or to the highest degree or extent. Used with many adjectives and adverbs to form the superlative degree: most honest; most impatiently.
2. Very: a most impressive piece of writing.
3. Informal. Almost: Most everyone agrees.
Idiom:
at (the) most
At the maximum: We saw him for ten minutes at the most. She ran two miles at most.
Note that there is nothing in there about a 'plurality'. Most means greatest in number, the greatest amount or the greatest part or number. Note that 20% of anything is not the greatest part or number of the whole, while it may be the largest piece, it is not the 'most'.
I wish people would be more precise in their use of language.
My journal has hot
So for example, if candidate A wins 45%, candidate B wins 40% and candidate C wins 15% that would be a plurality, not a majority, yet candidate A would have won the most votes. most is an imprecise word, it can mean a plurality or a majority, if you truly want to be precise in your language use, use plurality or majority instead of most. I take "greatest part of the whole" to mean that of all the individual parts, this is the largest. If you cut a loaf of bread into three unequal peices, one peice will be the greatest, i.e. larger than the other two individual peices taken seperately. Assuming you gave each peice to one person, the person who got the largest peice would then have the most bread. If two peices were given to one person, they may or may not now have the most bread, depending on the size of the individual peices.
Regardless of this, i think your confusion is actually coming from my use of the word "the" in my sentence. When i say"in order to be able to communicate with the most people, you would learn english" the most implies the maximum amount possible (while still only learning one language) Your contention that "most people dont know english" is correct, because your useage implies that most means a majority, and combines all the other languages into one (not english)
Whoops, you're absolutely right. I hate it when I lose sight of the forest for the trees.
At any rate, 50,000,000 people is still a lot. And the rest of my post stands.
End of lesson. You may press the button.
...though the Church of Latter Day Saints translates the Bible according to Joseph Smith.
That is incorrect. The Church of Jesus Christ of Latter Day Saints uses the King James Version of the Bible in English, and in foreign languages we use a commonly used version of the Bible in that language (such as the Reina Valera in Spanish).
TTFN