| Pages in topic: [1 2] > |
Will software engineers ever leave translators alone? Latest evil plan to destroy the profession Thread poster: Maria Eugenia Farre
|
|---|
Hi there,
Here\'s an interesting article about a new project that could have a huge impact in the profession. It\'s a kind of translation Napster where people would exchange strings for translation, for free, needless to say.
URL http://www.newscientist.com/news/news.jsp?id=ns99992115
Distributed program to translate many langu... See more Hi there,
Here\'s an interesting article about a new project that could have a huge impact in the profession. It\'s a kind of translation Napster where people would exchange strings for translation, for free, needless to say.
URL http://www.newscientist.com/news/news.jsp?id=ns99992115
Distributed program to translate many languages
16:47 02 April 02
NewScientist.com news service
A US software designer plans to harness the brains of the world\'s computer users to build a multilingual translation database. Brian McConnell believes it could provide a free way to translate the many languages not included in existing online translators.
McConnell is releasing a new distributed computer program, which works like programs such as SETI@home. However, while most distributed computer projects make use of spare computer power to perform complex computational tasks, in this case people will be asked to provide short translations themselves.
\"It\'s a clever twist on distributed computing,\" says McConnell. \"In this case the computers are people\'s brains.\"
The World Wide Lexicon (WWL) project will need multilingual volunteers to download a software program. This will automatically detect when the computer user is less busy and ask them to translate a word or phrase. However, some experts warn that the system may lack the quality of conventional dictionaries.
There are numerous online translation services for common languages, such as French, German and English, but McConnell says there are very few translators suited to less common languages. He hopes that people who speak more unusual languages will volunteer as translators.
Quality assurance
The first version of the program will be demonstrated at the O\'Reilly Emerging Technologies Conference in California in May 2002. But some experts are already unsure about the practicality of the system.
\"One of the main problems is quality assurance,\" says Ramesh Krishnamurthy, a linguistics expert at the University of Wolverhampton, in the UK. \"Translation is a highly developed skill.\"
McConnell concedes that this could be a problem and hopes to develop an automatic system for peer review, to ensure that translations are accurate.
But Paul Rayson, a research fellow at Lancaster University, adds that unskilled translators may confuse the meaning of individual words. \"The problem is you generally need the context to get a good translation,\" he says.
Spider web
To get the WWL database started, McConnell has designed a \"spider\" program to roam the web and select common words from foreign web sites. These will be sent to relevant volunteers for translation. Translations will be uploaded to one of a number of different servers, which will spread the information to other servers. The numerous servers should give the system more stability.
The next stage begins when a sufficiently large word database has been built. Users will then be able to download another client program and search these servers for different words. Words that are not found will be sent to volunteers for translation.
The WWL has been designed using the Simple Object Access Protocol (SOAP). McConnell says this should make it possible to integrate the client software into other computer applications. He hopes that other software developers will make use of the translation network, as an add-on to web browsers, for example.
Will Knight
This story is from NewScientist.com\'s news service - for more exclusive news and expert analysis every week subscribe to New Scientist print edition.
▲ Collapse
| | | |
| Here's my 2 Eurocents | May 23, 2002 |
Something like that is highly unlikely to provide satisfactory translation solutions, as it will be impossible to carry out the necessary level of Quality Assurance. Users will use such a system at their own risk and anyone who is aware of the complexity of the translation process will probably (hopefully ) have the sense to stay away from it.
Volunteers may include many unqualified individuals - many people seem to think that translation ... See more Something like that is highly unlikely to provide satisfactory translation solutions, as it will be impossible to carry out the necessary level of Quality Assurance. Users will use such a system at their own risk and anyone who is aware of the complexity of the translation process will probably (hopefully) have the sense to stay away from it.
Volunteers may include many unqualified individuals - many people seem to think that translation is easy and that it\'s enough just to have an A Level or even GCSE in the language in order to translate for a living ... well, I guess we know differently ... ▲ Collapse
| | | |
| Interesting, but there are a few fundamental failings... | May 24, 2002 |
1. It relys on people to willingly volunteer translation of phrases, and \"hopes to develop an automatic system for peer review\". There is no clear plan for finding people to volunteer these translations, or finding people to check them.
2. Context. The basis of this system is the translation of phrases, which can never truly understand context. Even if a translator translates the phrase correctly, that does not mean the same phrase will make sense in a different sente... See more 1. It relys on people to willingly volunteer translation of phrases, and \"hopes to develop an automatic system for peer review\". There is no clear plan for finding people to volunteer these translations, or finding people to check them.
2. Context. The basis of this system is the translation of phrases, which can never truly understand context. Even if a translator translates the phrase correctly, that does not mean the same phrase will make sense in a different sentence, paragraph, document, etc.
3. At the end of the day, it is still software. I recall my linguistics professor teaching one basic principle of human speech: every sentence is original and uses a different combination of words (unlike animal speech). For that reason alone, translation software based upon phrases or sentences will fail, simply because no-one can ever capture the limitless combinations of sentences and phrases that can be used.
While I do praise any effort to gather extensive vocabularies from around the world (which would be of great benefit to our profession), trying to turn that into a translation mechanism will fail miserably.
▲ Collapse
| | | |
| Don't worry... | May 24, 2002 |
... machines will NEVER be able to think.
Regards, 
BSD
| | |
|
|
|
| Look what a machine translation can do to a Beatles' song. | May 24, 2002 |
Translation from English into %#@!)*&^%$#
The long and winding road, that leads to your door,
Will never disappear, I\'ve seen that road before,
It always leads me here, lead me to your door.
The wild and windy night, that the rain washed away,
Has left a pool of tears, crying for the day.
Why leave me standing here? Let me know the way.
... See more Translation from English into %#@!)*&^%$#
The long and winding road, that leads to your door,
Will never disappear, I\'ve seen that road before,
It always leads me here, lead me to your door.
The wild and windy night, that the rain washed away,
Has left a pool of tears, crying for the day.
Why leave me standing here? Let me know the way.
Many times I\'ve been alone, and many times I\'ve cried.
Anyway you\'ll never know the many ways I\'ve tried,
And still they lead me back, to the long winding road.
You left me standing here, a long, long time ago.
Don\'t leave me waiting here, lead me to you door!
But still they lead me back, to the long and winding road.
You left me standing here, a long, long time ago.
Don\'t keep me waiting here, lead me to you door!
Yea, yea, yea, yea.
El camino largo y de la bobina, de que conduce a su puerta, nunca desaparecerá, yo ha visto ese camino antes, él me conduce siempre aquí, me conduce a su puerta. La noche salvaje y ventosa, de que que la lluvia se lavó lejos, ha dejado una piscina de rasgones, gritando para el día. ¿Por qué déjeme que está parado aquí? Déjeme saber la manera. Muchas veces he sido solo, y muchas veces que he gritado. De todas formas usted nunca sabrá las muchas maneras que he intentado, y todavía me conducen detrás, al camino largo de la bobina. Usted me dejó que estaba parado aquí, hace un rato largo, largo. ¡No me deje que espera aquí, me conducen a usted puerta! Pero todavía me conducen detrás, al camino largo y de la bobina. Usted me dejó que estaba parado aquí, hace un rato largo, largo. ¡No me guarde el esperar aquí, me conducen a usted puerta! Yea, yea, yea, yea.
▲ Collapse
| | | |
| Maria Eugenia, I asked the machine to "translate" the song into Portuguese. | May 24, 2002 |
Sorry for not realizing earlier. I had the translation into Spanish, but got this one specially for you.
~~~~~~~~~~~~~~~~~~~~~~~~
The long and winding road, that leads to your door,
Will never disappear, I\'ve seen that road before,
It always leads me here, lead me to your door.
The wild and windy night, that the rain washed away,
Has left a pool of tears, crying for the day.
Why leave me standing here? Let me know ... See more Sorry for not realizing earlier. I had the translation into Spanish, but got this one specially for you.
~~~~~~~~~~~~~~~~~~~~~~~~
The long and winding road, that leads to your door,
Will never disappear, I\'ve seen that road before,
It always leads me here, lead me to your door.
The wild and windy night, that the rain washed away,
Has left a pool of tears, crying for the day.
Why leave me standing here? Let me know the way.
Many times I\'ve been alone, and many times I\'ve cried.
Anyway you\'ll never know the many ways I\'ve tried,
And still they lead me back, to the long winding road.
You left me standing here, a long, long time ago.
Don\'t leave me waiting here, lead me to you door!
But still they lead me back, to the long and winding road.
You left me standing here, a long, long time ago.
Don\'t keep me waiting here, lead me to you door!
Yea, yea, yea, yea.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
A estrada longa e do enrolamento, aquela conduz a sua porta, nunca desaparecerá, mim viu essa estrada antes, ele conduz-me sempre aqui, conduz-me a sua porta. A noite selvagem e ventosa, aquela que a chuva se lavou afastado, deixou um pool dos rasgos, gritando para o dia. Por que me deixe que está aqui? Deixe-me saber a maneira. Muitas vezes eu fui sozinho, e muitas vezes que eu gritei. Em todo o caso você nunca saberá muitas maneiras que eu tentei, e ainda conduzem-me para trás, à estrada longa do enrolamento. Você deixou-me que está aqui, uma estadia longa, longa há. Não me deixe que espera aqui, conduzem-me lhe porta! Mas ainda conduzem-me para trás, à estrada longa e do enrolamento. Você deixou-me que está aqui, uma estadia longa, longa há. Não me mantenha esperar aqui, conduzem-me lhe porta! Yea, yea, yea, yea.
▲ Collapse
| | | |
Nathalie M. Girard, ALHC (X) 
English to French
+ ...
| Don't worry, it's not the first time... | May 24, 2002 |
... that someone comes up with plans in regards to quick translations. The results are laughable at the most.
Just look at the *human* input needed to even run CATs
I think that it\'s nice to dream up these things and that in the end, some progress may come out of it in the way we work and/or find information.
Either more people will realise how hard it is to obtain even the most basic ... See more ... that someone comes up with plans in regards to quick translations. The results are laughable at the most.
Just look at the *human* input needed to even run CATs
I think that it\'s nice to dream up these things and that in the end, some progress may come out of it in the way we work and/or find information.
Either more people will realise how hard it is to obtain even the most basic translation using that interface (if it ever comes alive)...which will them make them turn to us, or, the developpers will soon realise that they did not fully think this through because each language and dialect is so unique and *alive*.
It changes continuously...living language...
I do admit that it is irritating to see our profession put down as something easy to do. It\'s far from that, we know it and this means we\'ll need to continue educating our clients.
Best regards,
Nathalie
▲ Collapse
| | | |
| We will all meet at the beaches of Bali :-) | May 24, 2002 |
No worries about machine translations and Translation Napsters.
I don\'t have to write much about machine translations - just did a 100,000 term checking/translation job with ProZ colleagues for a Japanese/English/German machine translation program...no way that a machine can ever make the fine and detailled decision of our complicated and fuzzy thinking. The computer usually takes the wrong turn with a decision.
The Napster idea will not work either, a... See more No worries about machine translations and Translation Napsters.
I don\'t have to write much about machine translations - just did a 100,000 term checking/translation job with ProZ colleagues for a Japanese/English/German machine translation program...no way that a machine can ever make the fine and detailled decision of our complicated and fuzzy thinking. The computer usually takes the wrong turn with a decision.
The Napster idea will not work either, as Napster was based on a database of songs in each other\'s computer - exactly defined \"products\". I just imagine that I would exchange a Toyota translation with a Honda or Nissan translation of the same topic ... the client would hit the ceiling! They are all proud of their own specific terminology.
At least our generation of translators will not have sleepless nights because of computer translations.
So what\'s about the headline?
The Internet is a big leveler. In each and any translation related forum you will see complaints about \"sinking rates\", \"ridiculous pricing\", etc. Let\'s face it - we became a borderless profession and what is a perfectly normal translation fee in Argentina, Hungary or Romania, would not suffice to pay my office rent in Yokohama, the most expensive business city in the world.
But more and more of my clients move their business abroad, which is very easy due to the Internet. I can feel the pinch.
So I can see the writing on the wall that one day we will all be sitting in bamboo huts with palm leaf roofs and broadband connections at the beaches of Bali. Maybe not the worst of all results? ▲ Collapse
| | |
|
|
|
At least they are acknowledging something good...
\"Translation is a highly developed skill.\"
We should thank this guy for this... not everyone is willing to accept it...
And have to say that we should not worry, as other colleages have said, it is imposible for machines to discover the intricacies of the language and specially to resolve the diversity related with translations.
Sometimes, even the human transl... See more At least they are acknowledging something good...
\"Translation is a highly developed skill.\"
We should thank this guy for this... not everyone is willing to accept it...
And have to say that we should not worry, as other colleages have said, it is imposible for machines to discover the intricacies of the language and specially to resolve the diversity related with translations.
Sometimes, even the human translators have problems to resolve them!
Don\'t worry and be Happy! ▲ Collapse
| | | |
edsapir
Portuguese to English
+ ...
| Don't discount it out of hand. | May 24, 2002 |
I don\'t think it makes much sense to be threatened by this program.
Russell points out that it isn\'t clear how a peer review system will be implemented, and this is a key point. A pro is a pro, and people who need something translated right will call one.
As both Bertha and Nathalie point out, machine translation has nowhere near the accuracy necessary to unseat trained professionals like yourselves. That translation of the Beatles sure doesn\'t scan. ... See more I don\'t think it makes much sense to be threatened by this program.
Russell points out that it isn\'t clear how a peer review system will be implemented, and this is a key point. A pro is a pro, and people who need something translated right will call one.
As both Bertha and Nathalie point out, machine translation has nowhere near the accuracy necessary to unseat trained professionals like yourselves. That translation of the Beatles sure doesn\'t scan. 
So, if McConnell\'s system isn\'t really a threat to translators, who can make use of it?
Perhaps speakers of less common languages, who do not have the benefit of a large body of professional translators working to develop material for the speakers.
Suppose all of the speakers of, say, Ojibwa decided to contribute to this system. That speech community, and indeed, that culture, could benefit from being networked. Is it fair to reject the technology on other grounds?
The issue of fair payment for translators is an important one, and will have to be resolved like it is in every profession (as many professions must adapt to the internet), but that issue is a separate one from the development of new technology.
The internet has multiplied the amount of translation needs to be done, not decreased it.
If technology reaches a point where realistic translation is feasible, then we\'ll have far greater issues to deal with than how it affects one profession. That would mean we\'d be arguing about what it means to be human, and other wacky stuff!
This program is just a little piece in a big, big puzzle. Looks like fun to me.
Cheers,
-Patrick
~
fieldmethods.net
\"All the News that\'s Fit to Parse\" ▲ Collapse
| | | |
Mats Wiman
Sweden
Local time: 02:18
Member (2000)
German to Swedish
+ ...
In memoriam | The beaches of Bali are too expensive | May 24, 2002 |
I agree with you Hans-Henning, only Bali is too expensive .
Translators will be leaders in finding shore sites with a broadband connection.
I\'ve got mine in North Sweden although there are cheaper alternatives.
| | | |
| I can't wait for it | May 24, 2002 |
as it would be just another powerful(?) tool for us translators. I wonder if paticipants to such a system would be granted Kudoz points, too.
To extinguish our profession it will take more than a smart interconnection of todays computers.
Anyway, given the actual pace of computer evolution we can\'t be sure until the end of time...
| | |
|
|
|
| Thanks for your enlightened comments | May 24, 2002 |
I totally agree with you.
I don\'t understand where they are going to get the volunteers either, and I am 100% per cent with you when you mention the context problem.
How can anyone translate without context?
Gee...
What annoys me is that scientists keep dreaming of the Universal Translator, as featured in the Enterprise. And they try so hard, in so many different ways, that I am afraid one day they are going to ... See more I totally agree with you.
I don\'t understand where they are going to get the volunteers either, and I am 100% per cent with you when you mention the context problem.
How can anyone translate without context?
Gee...
What annoys me is that scientists keep dreaming of the Universal Translator, as featured in the Enterprise. And they try so hard, in so many different ways, that I am afraid one day they are going to get it right.
No matter how clumsy they still are Automatic Translators are getting better. And what is really scary about this project is that they are using human translators, although unskilled.
Wouldn\'t that be the equivalent of a worldwide mass dumping of a profession?
ME
Quote:
On 2002-05-24 00:24, rgillis wrote:
1. It relys on people to willingly volunteer translation of phrases, and \"hopes to develop an automatic system for peer review\". There is no clear plan for finding people to volunteer these translations, or finding people to check them.
2. Context. The basis of this system is the translation of phrases, which can never truly understand context. Even if a translator translates the phrase correctly, that does not mean the same phrase will make sense in a different sentence, paragraph, document, etc.
/snip/
▲ Collapse
| | | |
Quote:
On 2002-05-24 01:39, CTIS wrote:
/snip/
I do admit that it is irritating to see our profession put down as something easy to do. It\'s far from that, we know it and this means we\'ll need to continue educating our clients.
Nathalie
Hi Nathalie,
My thoughts precisely. Nobody would dare think up of a Legal Advice N... See more Quote:
On 2002-05-24 01:39, CTIS wrote:
/snip/
I do admit that it is irritating to see our profession put down as something easy to do. It\'s far from that, we know it and this means we\'ll need to continue educating our clients.
Nathalie
Hi Nathalie,
My thoughts precisely. Nobody would dare think up of a Legal Advice Napster or a Medical Diagnostic Napster, firstly because no lawyer in his/her right mind would volunteer for such a preposterous and ill-fated project. And if they managed to find lay volunteers willing to provide legal advice or mumbo-jumbo diagnostics, no one would rely on them. But people who aren\'t aware of the intrincacies of translation are still willing to buy and use automatic translators and other tools. And worse, many people with poor knowledge of languages would be willing to volunteer for the Translation Napster, in my opinion. Kind of a show-off I guess...When are people going to learn that there is an immense commitment of time and effort until you can call yourself a translator ?
ME ▲ Collapse
| | | |
| Less common languages: I doubt it! | May 24, 2002 |
Quote:
On 2002-05-24 05:49, edsapir wrote:
So, if McConnell\'s system isn\'t really a threat to translators, who can make use of it?
Perhaps speakers of less common languages, who do not have the benefit of a large body of professional translators working to develop material for the speakers.
Suppose all of the speakers of, say, Ojibwa decided to contribute to this system. That speech comm
... See more Quote:
On 2002-05-24 05:49, edsapir wrote:
So, if McConnell\'s system isn\'t really a threat to translators, who can make use of it?
Perhaps speakers of less common languages, who do not have the benefit of a large body of professional translators working to develop material for the speakers.
Suppose all of the speakers of, say, Ojibwa decided to contribute to this system. That speech community, and indeed, that culture, could benefit from being networked. Is it fair to reject the technology on other grounds?
/snip/
-Patrick
~
fieldmethods.net
\"All the News that\'s Fit to Parse\"
Hi Pat,
Unfortunately I doubt that Ojibwa speakers or other lesser spoken languages have that many Internet connections available to participate in the project. The digital divide is real whether we like it or not (and I certainly don\'t). Contribution from these language groups would be minimal. What I really think is that McConnell is trying to sugarcoat the bitter pill by saying \"look, it\'s a humanitarian project\". BS, IMHO. He is already speaking of integration as a browser plug-in, and that could eventually take a bite of the Web-site localization market (provided that he gets the peer reviews, and clever volunteers, and the thing doesn\'t crash your puter, many factors there, I am speaking hipothetically). You know that the people who are going to sign up are the speakers from the more affluent/connected countries.
ME ▲ Collapse
| | | |
| Pages in topic: [1 2] > |
Login to reply/comment 
Translation news
