|
 |
Learning completely offline/analog | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6600 days ago 9753 posts - 15779 votes  4 sounds Speaks: Russian*, English, FinnishC1, Latin, German, Italian, Spanish, Portuguese Studies: Danish, Romanian, Polish, Belarusian, Ukrainian, Croatian, Slovenian, Catalan, Czech, Galician, Dutch, Swedish | Message 25 of 34 24 October 2014 at 7:33pm | IP Logged | |||||||||||
|
Sorry if I'm stating the obvious but as of now pairs like Maltese-Swahili are translated via English. I'm not very optimistic about the pairs not involving English, tbh.
1 person has voted this message useful | ||||||||||||
| Ari Heptaglot Senior Member Norway Joined 6585 days ago 2314 posts - 5695 votes Speaks: Swedish*, English, French, Spanish, Portuguese, Mandarin, Cantonese Studies: Czech, Latin, German | Message 26 of 34 24 October 2014 at 7:44pm | IP Logged | |||||||||||
Then there's good hope for the major European languages! Surely the EU is pumping out enormous amounts of parallel texts in its main languages? And with the steady growth of bureaucracy, the corpus growth should be exponential! 1 person has voted this message useful | ||||||||||||
 emk Diglot Moderator United States Joined 5535 days ago 2615 posts - 8806 votes Speaks: English*, FrenchB2 Studies: Spanish, Ancient Egyptian Personal Language Map | Message 27 of 34 24 October 2014 at 8:27pm | IP Logged | |||||||||||
Those parallel texts from the UN, EU and Canadian government are actually a key resource for machine translation. That's one reason why Google can translate the front page of Le Monde, but barfs horribly on the more idiomatic articles on a site like Topito. Here's a paragraph from Le Monde, showing what happens when you have a good parallel corpus based on governmental and corporate documents:
This is broken but usable. But if you want to see massive translation failure, try this article from Topito (warning, cruelty to seagulls):
The underlying data set makes a huge difference, as you can see. 1 person has voted this message useful | ||||||||||||
| robarb Nonaglot Senior Member United States languagenpluson Joined 5062 days ago 361 posts - 921 votes Speaks: Portuguese, English*, German, Italian, Spanish, Dutch, Swedish, Esperanto, French Studies: Mandarin, Danish, Russian, Norwegian, Cantonese, Japanese, Korean, Polish, Greek, Latin, Nepali, Modern Hebrew | Message 28 of 34 24 October 2014 at 8:49pm | IP Logged | |||||||||||
Right, do you see how that's using a million-dollar, immobile, hazardous machine to transmit "yes" and "no"? It's cool, and it will become more powerful, but it's not going to replace language any time soon. It's most promising for those who can't move their body, for whom there are few other options. As for EEG, the point was that EEG is the most powerful known brain-scanning technique that is portable, safe, affordable and doesn't require implanting anything in the head. It is incapable of supporting the kind of communication you mentioned, and there isn't another option. There is no fMRI helmet, any more than there is a hydroelectric power plant on an airplane. 1 person has voted this message useful | ||||||||||||
| Ari Heptaglot Senior Member Norway Joined 6585 days ago 2314 posts - 5695 votes Speaks: Swedish*, English, French, Spanish, Portuguese, Mandarin, Cantonese Studies: Czech, Latin, German | Message 29 of 34 24 October 2014 at 9:05pm | IP Logged | |||||||||||
Well, you can do a lot more than just yes and no with much cheaper and lighter machines. This video is four years old, and it shows a machine for a few hundred dollars that allows you to push, pull, rotate etc. a virtual object with your mind. I'm sure research has improved since then. Still, of course, actually reading your mind and feelings and, even weirder, writing those feelings to another brain is very, very far off. Whereas babel fish programs combining speech recognition, translation and TTS is already here, albeit clumsy and problematic. 4 persons have voted this message useful | ||||||||||||
| Cavesa Triglot Senior Member Czech Republic Joined 5012 days ago 3277 posts - 6779 votes Speaks: Czech*, FrenchC2, EnglishC1 Studies: Spanish, German, Italian | Message 30 of 34 24 October 2014 at 9:32pm | IP Logged | |||||||||||
|
The point of the EEG is that it is not exactly a brain scanning technique, unlike the MRI. By the nature of the method, it will never be possible to get any results similar to the MRI. But there are more methods than just the MRI. However, the point is that a millions dollars costing machine can have a cheap and practical variant in future as it has happened for so many things in the past. And if it wasn't already clear, I think both the thought interpreting tools and machine speech translator are things not needed, too expensive to make and potentially very harmful.
If everything is getting machine translated through English, as Sertent says, than there is no point in machine translation at all. The world is learning English, even older people are learning, most people likely to ever need English are learning it. So why make a X-En-Y machine translation, when the X and Y speakers can just use English? Waste of money on researching such a tool, in my opinion. Reading the feelings and some kinds of thoughts is not that far off. The world of marketing is putting lots of money into such researches, for their purposes. What do you want, what do you think about the scene before you, those are the first questions, just the beginning. Really, I think machine translation is actually just as potentially dangerous and practically stupid idea as mind reading, that was my point, not that we should read thoughts. Well, people have always complained about the humankind getting dumber but recent researches (published two years ago or so) actually hinted the IQ of the population has been declining for a long time. The stupid no longer die of it, most people don't need to use their potential to the fullest as our lives have become very comfortable. 1 person has voted this message useful | ||||||||||||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6600 days ago 9753 posts - 15779 votes 4 sounds Speaks: Russian*, English, FinnishC1, Latin, German, Italian, Spanish, Portuguese Studies: Danish, Romanian, Polish, Belarusian, Ukrainian, Croatian, Slovenian, Catalan, Czech, Galician, Dutch, Swedish | Message 31 of 34 24 October 2014 at 10:15pm | IP Logged | |||||||||||
|
I was referring to google or likely any service that offers exotic combinations. Afaiu some Russian software can translate directly to and from French, Spanish, German, Ukrainian, without involving English. There are likely other options for this or that specific combination, too.
Anyway, sadly there are still tons of people using inefficient methods and not getting anywhere. Edited by Serpent on 24 October 2014 at 10:17pm 1 person has voted this message useful | ||||||||||||
| robarb Nonaglot Senior Member United States languagenpluson Joined 5062 days ago 361 posts - 921 votes Speaks: Portuguese, English*, German, Italian, Spanish, Dutch, Swedish, Esperanto, French Studies: Mandarin, Danish, Russian, Norwegian, Cantonese, Japanese, Korean, Polish, Greek, Latin, Nepali, Modern Hebrew | Message 32 of 34 25 October 2014 at 4:06am | IP Logged | |||||||||||
|
If you want to translate from Danish to Norwegian, you're probably better off leaving it alone than going through
English ;) OK, but maybe we can take this back to technology and language learning. If you had a good machine translator, what would you do with it? You could create your own parallel texts. You could also use it to help decipher something you didn't get while reading. But parallel texts made by humans already exist, and teachers or forums can help decipher a confusing bit of text. So it wouldn't really be a huge advance, just another tool to make learning more accessible and convenient under certain circumstances. Perhaps a stronger future contribution of science/technology would be a deeper understanding of how we learn. We know some things now, like some factors that impact retention in flashcard study, what predicts student success in the classroom, sleep and memory, etc. None of this stuff has a really huge impact, though. Maybe in some distant future we won't have people complaining "I don't know what's wrong, but I'm not learning," because we'll understand the underlying processes in great detail, not just some external factors that influence their outcomes. 1 person has voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
