| Pages in topic: < [1 2 3 4 5 6 7 8 9] > | DeepL Thread poster: Michael Mangelsdorf
|
|---|
CafeTran Training (X)
Netherlands
Local time: 14:18
| Term Finding Tool | Aug 31, 2017 |
Claudio Porcellana wrote:
my opinion is that this tool is very reliable to find "pieces of a translation", less commonly a whole good translation, but a thinking brain is still needed (luckily enough) to properly join these pieces together, so I think we will not replaced in the short by HAL 9000
I agree. I think that DeepL is (another) useful "term suggestion tool". It doesn't present ready-made translations to the professional translator, but it can remind him of certain alternative terms (that he already knows, but that didn't come to his mind quickly enough) or even put him on track of a correct term (that he didn't think of himself and that he should further investigate and verify).
| | | | Claudio Porcellana (X) 
Italy
Local time: 14:18
English to Italian
| Term Finding Tool | Aug 31, 2017 |
I can add that DeepL is particularly able, as opposite to GT, to find stylish locutions, so I find it very useful when the exact concept is already in my mind, but the exact writing is still on the tip of my tongue
[/quote]
I agree. I think that DeepL is (another) useful "term suggestion tool". It doesn't present ready-made translations to the professional translator, but it can remind him of cert... See more I can add that DeepL is particularly able, as opposite to GT, to find stylish locutions, so I find it very useful when the exact concept is already in my mind, but the exact writing is still on the tip of my tongue
[/quote]
I agree. I think that DeepL is (another) useful "term suggestion tool". It doesn't present ready-made translations to the professional translator, but it can remind him of certain alternative terms (that he already knows, but that didn't come to his mind quickly enough) or even put him on track of a correct term (that he didn't think of himself and that he should further investigate and verify). [/quote] ▲ Collapse
| | | | Michael Beijer
United Kingdom
Local time: 13:18
Member (2009)
Dutch to English
+ ...
CafeTran Training wrote:
Claudio Porcellana wrote:
my opinion is that this tool is very reliable to find "pieces of a translation", less commonly a whole good translation, but a thinking brain is still needed (luckily enough) to properly join these pieces together, so I think we will not replaced in the short by HAL 9000
I agree. I think that DeepL is (another) useful "term suggestion tool". It doesn't present ready-made translations to the professional translator, but it can remind him of certain alternative terms (that he already knows, but that didn't come to his mind quickly enough) or even put him on track of a correct term (that he didn't think of himself and that he should further investigate and verify).
… once the API is ready, it may be possible to put it to very good use as a sub-segment lookup/auto-assembling tool in a (sufficiently clever) CAT tool.
| | | | CafeTran Training (X)
Netherlands
Local time: 14:18
| Interesting video explaining the basics (in German) | Sep 1, 2017 |
Mr Frahling comes across as a very sympathetic person. In this video he presents the principles of DeepL: https://youtu.be/DrOS6wyEzFI
The training of the engine has just started.
@Michael: Neurons that match with native speaking/non-natives.
[Edited at 2017-09-01 06:10 GMT]
| | |
|
|
|
| Convolutional Neural Networks | Sep 1, 2017 |
Journalists are asking for compulsory labelling for computer-generated articles. DeepL in a few months may do what Google AlphaGo did to the game of Go with related algorithms i.e. own the game forever. Would it help to ask for compulsory labelling of computer generated translations? Will the consumer pay a premium for those to let us live? So many different fields, same questions : )
New York times March 7 2015:
A shocking amount of what we’re reading is created not by huma... See more Journalists are asking for compulsory labelling for computer-generated articles. DeepL in a few months may do what Google AlphaGo did to the game of Go with related algorithms i.e. own the game forever. Would it help to ask for compulsory labelling of computer generated translations? Will the consumer pay a premium for those to let us live? So many different fields, same questions : )
New York times March 7 2015:
A shocking amount of what we’re reading is created not by humans, but by computer algorithms. Can you tell the difference? Take the quiz.
https://www.nytimes.com/interactive/2015/03/08/opinion/sunday/algorithm-human-quiz.html?mcubz=3 ▲ Collapse
| | | | kakapo77
France
Local time: 14:18
French to Italian
+ ...
| Dramatic change | Sep 1, 2017 |
I think that the market of translation will dramatically shrink in a short time. The only field where humans should keep is literary texts, or highly specialized stuff such as phylosophy.
Then you guys can keep on saying that "DeepL does not work well on lists" or that "It is just another tool"... But basically, when you say that, you are looking for potential niches to survive. Which means that you understand that a big chunk of the market is bound of disappear -- and already has,... See more I think that the market of translation will dramatically shrink in a short time. The only field where humans should keep is literary texts, or highly specialized stuff such as phylosophy.
Then you guys can keep on saying that "DeepL does not work well on lists" or that "It is just another tool"... But basically, when you say that, you are looking for potential niches to survive. Which means that you understand that a big chunk of the market is bound of disappear -- and already has, indeed.
Now, if the market shrinks, then the law of supply and demand will do the rest. You also have to consider that this piece of software will likely be improved in the future, and that other ones will appear.
I think that this is sad and eventually even risky for humanity, but there is no point in denying the reality. ▲ Collapse
| | | | neilmac
Spain
Local time: 14:18
Spanish to English
+ ...
I haven't tried it yet, but if it has been developed by Linguee, I will be as wary about the results as I am every time I consult that particular site. I'm not saying that it isn't helpful, but also an occasional great source of amusement due to some of the woeful translations it offers. When they learn to sort out the wheat from the chaff, I'll feel more confident about using this type of "tool"…
PS: I just tried it out. I dropped a chunk of 300 words into it, a list of items fro... See more I haven't tried it yet, but if it has been developed by Linguee, I will be as wary about the results as I am every time I consult that particular site. I'm not saying that it isn't helpful, but also an occasional great source of amusement due to some of the woeful translations it offers. When they learn to sort out the wheat from the chaff, I'll feel more confident about using this type of "tool"…
PS: I just tried it out. I dropped a chunk of 300 words into it, a list of items from a translation I'm currently working on. The result seems not bad, at least on a par with GT, and it was very quick, almost instant.
[Edited at 2017-09-01 08:02 GMT] ▲ Collapse
| | | | Tom in London
United Kingdom
Local time: 13:18
Member (2008)
Italian to English
| DeepL translates, but does not understand. | Sep 1, 2017 |
I've just been trying it out on a translation I'm working on at the moment. Like Lingee or GT, DeepL helps a little (mainly because it is impersonal, and for that reason may sometimes suggest different, more refreshing words than the ones I might have chosen out of habit).
But in my language pair (Italian to English) DeepL doesn't understand the different ways of constructing sentences in these two languages, with the result that a perfectly acceptable Italian sentence construction ... See more I've just been trying it out on a translation I'm working on at the moment. Like Lingee or GT, DeepL helps a little (mainly because it is impersonal, and for that reason may sometimes suggest different, more refreshing words than the ones I might have chosen out of habit).
But in my language pair (Italian to English) DeepL doesn't understand the different ways of constructing sentences in these two languages, with the result that a perfectly acceptable Italian sentence construction is translated mechanically into the same construction in English, where it sounds extremely strange and would not be acceptable.
So in the case of the particular translation I'm doing today, which is very typical of the kind of work I do, Deepl is able to mechanically translate a paragraph, but then I have to completely rejig that paragraph to make it flow, and in most cases to use better words than those that were chosen (mechanically guessed at) by DeepL. The point is: MT *translates* but doesn't *understand*.
The owners of DeepL would like me to input those better words that I have chosen, because they are going to be relying on us hard-working translators inputting corrections, suggesting better words, etc, free of charge.
That's the Achilles heel of this business, which has got nothing to do with computers, algorithms, and IT, but good old human effort.
For obvious reasons, therefore, I would exhort colleagues not to offer any kind of collaboration to the owners of DeepL unless DeepL is willing to pay for this service. That would include not inputting corrections to DeepL.
[Edited at 2017-09-01 08:38 GMT] ▲ Collapse
| | |
|
|
|
Oksana Weiss
Germany
Local time: 14:18
Member (2011)
English to Russian
+ ...
DeepL does not offer Ukrainian and Russian, among others. So no need to be concerned for those working with these languages. Basically, today it offers only 7 languages, which is not that much compared to GT. For other languages it remains to be seen...
| | | | Rebecca Davis
United Kingdom
Local time: 13:18
French to English
+ ...
... that is rarely mentioned is where the users of MT applications like Deepl think they are going to get high-quality editors from. In my (very limited) experience, editing MT output is actually more time-consuming than translating it from scratch, mostly due to syntax and structural issues (as Tom points out), although extremely proficient editors will still be needed to avoid critical mistakes.
So who's up for this, presumably at a knock-down rate?
| | | | Tom in London
United Kingdom
Local time: 13:18
Member (2008)
Italian to English
Rebecca Davis wrote:
... that is rarely mentioned is where the users of MT applications like Deepl think they are going to get high-quality editors from. In my (very limited) experience, editing MT output is actually more time-consuming than translating it from scratch, mostly due to syntax and structural issues (as Tom points out), although extremely proficient editors will still be needed to avoid critical mistakes.
So who's up for this, presumably at a knock-down rate?
Precisely, Rebecca. You've hit the nail on the head or, translating that into Italian using Deepl "Hai colpito il chiodo sulla testa"
| | | | Claudio Porcellana (X)
Italy
Local time: 14:18
English to Italian
| editing MT output is actually more time-consuming than translating it from scratch | Sep 1, 2017 |
sure, but only a mad PM can ask you to do that!
PEMT is totally another thing than remaking the whole line
now, one year ago I did some PEMT audits for a big LP (to understand what is involved) and here is what I understood
Just in case you will be asked to do some PEMT
... See more sure, but only a mad PM can ask you to do that!
PEMT is totally another thing than remaking the whole line
now, one year ago I did some PEMT audits for a big LP (to understand what is involved) and here is what I understood
Just in case you will be asked to do some PEMT
Claudio
[Edited at 2017-09-01 10:59 GMT] ▲ Collapse
| | |
|
|
|
| If you built your career... | Sep 1, 2017 |
... translating the likes of consumer electronics manuals, EULAs or software interfaces, you have all the right to be worried, and, in fact, should have started worrying a while ago. Get working hard on moving upmarket, if you can.
The bottom line, in my experience: if there is something important at stake for the client, they are willing to pay well to protect it. The DeepLs of the internet may shrink the market and cut off its lower end, which would be very welcome.
I... See more ... translating the likes of consumer electronics manuals, EULAs or software interfaces, you have all the right to be worried, and, in fact, should have started worrying a while ago. Get working hard on moving upmarket, if you can.
The bottom line, in my experience: if there is something important at stake for the client, they are willing to pay well to protect it. The DeepLs of the internet may shrink the market and cut off its lower end, which would be very welcome.
I have no doubt that once it arrives for Slovak, it will be better than GT, because we have nowhere to move but up
[Edited at 2017-09-01 13:30 GMT] ▲ Collapse
| | | | MK2010
United States
Local time: 08:18
French to English
+ ...
The concern I would have, if this is done by the good folks over at Linguee, is that their French terms and expressions are about 80% French Canadian. For those of us translating from/into Parisian French, many of the terms our searches produce on the Linguee site are useless unless confirmed by another source (French or international body). I don't know if that will be the case with their MT, I doubt it, but if so, translators in this language pair need to be cautious.
| | | | Tom in London
United Kingdom
Local time: 13:18
Member (2008)
Italian to English
Similarly, they assume that English means American English, which makes DeepL inconvenient for use anywhere outside North America.
| | | | | Pages in topic: < [1 2 3 4 5 6 7 8 9] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » DeepL | CafeTran Espresso | You've never met a CAT tool this clever!
Translate faster & easier, using a sophisticated CAT tool built by a translator / developer.
Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools.
Download and start using CafeTran Espresso -- for free
Buy now! » |
| | TM-Town | Manage your TMs and Terms ... and boost your translation business
Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
