Poll: What is your take on neural machine translation? Thread poster: ProZ.com Staff
|
|---|
This forum topic is for the discussion of the poll question "What is your take on neural machine translation?".
View the poll results »
| | | |
| I will not use it in any way | Feb 13, 2019 |
I don’t use machine translation in any way whatever adjectives are added to the term… I fell in love with translation many moons ago and I'm hoping that by keeping my brain active, I will keep Alzheimer's at bay!
| | | |
neilmac
Spain
Local time: 07:29
Spanish to English
+ ...
As Google is apparently using NMT technology now, and I use Google, I suppose I'm already using it. However, I am usually suspicious of the people or businesses pushing this kind of thing as a panacea, as they're usually simply aiming to make a killing.
| | | |
I am always ready to learn about and accept new technology.
| | |
|
|
|
Ricki Farn
Germany
Local time: 07:29
English to German
| I will not use it professionally | Feb 13, 2019 |
... but to get an idea of what other people are saying on Social Media, it's sometimes better than nothing.
| | | |
Michael Harris 
Germany
Local time: 07:29
Member (2006)
German to English
| Do not know what it is | Feb 13, 2019 |
and judging by the results of the poll, most others as well.
I do not have the time to look this up and cannot be bothered either.
Why can´t anonymous give us an explanation?
| | | |
Frank Zou
China
Local time: 14:29
Member (2016)
Chinese to English
+ ...
From what I read, NMT is much more useful than the previous machine translation as it operates through big data. I'm myself a strong supporter of Human Translation, but it was until recently that I found out MT was very helpful in some cases. I think we should keep an open mind and get to know such technology before we are phased out without knowing it.
| | | |
Muriel Vasconcellos (X)
United States
Local time: 22:29
Spanish to English
+ ...
From what I've seen, it's a royal mess. Certainly not for use in serious translation because phrases that "sound good" may be quite different from the original meaning. It takes more time to be editing the output while constantly checking the input than it would take to translate from scratch.
Several of the new MT systems seem to be OK in the languages I know for "information-only" purposes. Neural MT "sounds" more like natural text, but I'm already seeing the kinds of trouble it c... See more From what I've seen, it's a royal mess. Certainly not for use in serious translation because phrases that "sound good" may be quite different from the original meaning. It takes more time to be editing the output while constantly checking the input than it would take to translate from scratch.
Several of the new MT systems seem to be OK in the languages I know for "information-only" purposes. Neural MT "sounds" more like natural text, but I'm already seeing the kinds of trouble it can lead to.
P.S. I've been working with MT output since 1980 and I haven't seen a big improvement in the language combination that I am most exposed to (Spanish into English). The more big data that's squished into the system I'm most familiar with, the more unusable it's getting to be.
[Edited at 2019-02-14 06:54 GMT] ▲ Collapse
| | |
|
|
|
Mario Freitas
Brazil
Local time: 03:29
Member (2014)
English to Portuguese
+ ...
| Neural is your brain not a machine | Feb 14, 2019 |
Muriel Vasconcellos wrote:
...Certainly not for use in serious translation
...Several of the new MT systems seem to be OK in the languages I know for "information-only" purposes.
Ricki Farn wrote:
...I will not use it professionally.
Just like MT, neural T is not intended for professional translators to use at work. There is a trend that way, of course, because it makes it easier, but it does not make it better whatsoever, quite the opposite.
[Edited at 2019-02-14 15:27 GMT]
| | | |
Login to reply/comment 
