Copier Auto-Translates Japanese to English 244
StCredZero writes "Wild. Fuji has created a photocopier that automatically translates documents from Japanese to English. That's pretty nuts. Apparently, the copier can figure out what sections are text, OCR the text, send it to a translation engine, and put the english back into place."
Re:Manga and Anime (Score:3, Interesting)
This OCR based stuff would still be handy for automatically translating manga I suppose though.
I know that there are a few things out there in Japanese that haven't been released in English yet I wouldn't mind.
Engrish (Score:3, Interesting)
Re:I just photocopied this article (Score:2, Interesting)
Re:It's impossible... (Score:5, Interesting)
Disclaimer: I'm a second language acquisition researcher and assessor.
I concur. Absolutely. Language is not pure information; it's information shorthand. It assumes a high degree of already-shared knowledge about the world. Some of these assumptions are near-universal; many are not.
Japanese and English (my languages) offer a great example, especially as it pertains to machine translation. Whereas English is a subject-predicate language, where basically all the information is encoded in the language stream, Japanese is a topic-comment language, where, once set, the "subject" is not re-stated until it changes. Beginning Anglophone learners of Japanese make the mistake of putting a "wa" to denote what they think of as the subject in every sentence, when it does not need to be there. "Wa" is a topic marker; not a subject marker.
This is a fundamentally different way of thinking about language and, therefore, about the world. Germanic languages seek to operate regardless of context; Asian languages seek to augment (or "comment on") it. If you've ever felt that Japanese people who speak English are beating around the bush or being vague, part of that is cultural, but part of that is the language of the culture that does not require explicitness. A big part of learning Japanese or, for Japanese people, of learning English is learning how to think about the world and about human interactions in a very different way.
Machines aren't human. They are information processors. They don't know what a "cat" is; they just know that it's a piece of code that can be slotted into a certain place in a set of syntax. Until machines are really intelligent (and I don't think that will be anytime soon), expect more crappy translation than useful. Anyone who tells you otherwise is probably selling something (a crappy machine translator, to be exact!).
Re:It's impossible... (Score:2, Interesting)
That's the main reason why I believe western dubs of Japanese anime suck so much. Japanese has very, very different "conventions", which pushes storyboard writers towards building up scenes that sound very awkward in English or any other western language, even with a decent enough translation and good voice acting.
Re:I just photocopied this article (Score:2, Interesting)
Granted, Jeff Dean first brought up Google Translate by pointing out how much machine translation sucks in general, so nobody's under the illusion we'll have reliable online translaters within a few months. However, there are a lot of intriguing and innovative ideas out there that are still being implemented, and we could have borderline-acceptable machine translation in the not-too-distant future. Human translators are not in danger.
Says someone who's never translated something. (Score:5, Interesting)
Yeah, 'cause researchers have long promised us that AI will reach us in 10 years. <sarcasm>
Seriously, I think you underestimate the difficulty of translating. Have you done any major foreign-language translation -- especially of conversational speech? My experience has primarily been with Japanese and English, and I'll tell you right now that it can be nightmarish.
Sentence fragments are the worst part. Japanese has a completely different word order from English. All modifiers (including phrases and clauses) come before the word they modify, and the language has a Subject-Object-Verb order. "I just saw the man who stole my friend's watch last Tuesday" becomes "Just I Last Tuesday friend's watch stole man saw." Now try translating that from Japanese to English when the sentence is cut in half.
Worse, the language has very different levels of allowed vagueness. "Complete" sentences in Japanese can contain just a descriptor or an action without any specification of who did/was what. Conversely, translating "3 of them" in English to Japanese is hard because you have to know "3 of what?" to know what counting suffix to use.
Another problem is that many very different words sound exactly the same when conjugated to the gerund or perfective forms. English has a number of homonyms, but there are MANY more opportunities for mix-ups if you don't have access to kanji to tell the semantic meaning apart because Japanese has a much more limited range of phonemes. For example, take "katte" which is the gerund form of the verbs "kau" (buy), "kau" (keep/raise), "karu" (cut), "karu" (spur on), and "katsu" (win). That's 5 completely different verbs that conjugate to the same sound. If they're written phonetically or your going from speech, then you have to be able to understand the meaning behind the words to translate. (Did I mention earlier that you may not have an explicit subject and object to go off of?)
Then you get into issues of translating things like politeness levels, different ways of addressing people, and other concepts that don't translate well into English or concepts like singular vs. plural that are dropped in going to Japanese. Let's not even consider puns and poetry!
These are not trivial issues. An automatic translator would need to somehow be able to conceptualize what a person is trying to speak about, which would require understanding the story being told and an ability to predict where they are going with it. This will require strong AI.
Accurate and intelligible translation is an art -- not a science -- because it requires an intuitive and empathetic ability to understand the mind of the speaker well enough to map their thoughts into a different method of expression.
Re:Manga and Anime (Score:3, Interesting)
Re:Great! (Score:2, Interesting)
I'm know I've used online translation services when I absolutely needed to know what something said and there was no native speaker around. In fact, since native speakers are generally unsure of exact translations (despite speaking both languages fluently) I find the poor translations online better than 'well, it means this, but there's this connotation and...'
Re:Says someone who's never translated something. (Score:2, Interesting)
You may well have an amazing statistical technique that can compare your web page sentence-by-sentence to a massive corpus of bilingual pages, but how is it going to know whether "kondo ha simasu" on your page means "this time I'll do it", "this time you'll do it", "next time they'll do it", "next time we'll do it", or any of a dozen more possible combinations? It can't, unless it is capable of understanding the context and knowing what the sentence refers to. Statistics just won't cut it.
You may say that this will nonetheless give a good enough basis for a human to go through and edit the output into a coherent English whole. And there is some truth in this, though the human will still need to be fluent in both Japanese and English and to refer to the original text frequently, so you won't save much money that way. But given that machine translation cannot, and will not for the foreseeable future, be able to stand alone without human intervention, why is this being done in a copier? I could just about understand a scanner, that would put the translation into a nice Word document for the said human to check and fix up before printing...
Not only into English (Score:5, Interesting)
Obviously the translator was all at sea.
Re:Says someone who's never translated something. (Score:3, Interesting)
What the grandparent post is saying is that you might not have a full sentence, nor might a full sentence even provide sufficient information to perform the translation accurately. If Japanese and English were so easy to translate back and forth, don't you think that humans [engrish.com] would have an easier time of it?
The method of analyzing bilingual pages is great for languages that have a similar structure to one another, like italian and spanish. In fact, this is part of how I learned spanish, by reading a bilingual copy of Don Quixote in high school. Slightly botching a translation from Spanish to English or vice versa might so much as raise an eyebrow briefly, but translating between Japanese and English requires an intimate knowledge of both languages. I do think that it will be possible with natural language processing systems in the near future (think less than a decade) but, no question, we're not that close right now. These guys saying that the technology won't be here within their lifetime have to be ancient or just forgetting how rapidly the pace at which technology accelerates has been increasing of late. How long ago was it that this here "Internet" only had a few hundred nodes?
Re:Says someone who's never translated something. (Score:2, Interesting)
I think you are missing the point a bit. A native Japanese speaker will not have any problem understanding the japanese sentence even with the insufficient context(from the perspective of say an Englishman) but he will have problem with a sentence of similar sort which is in English even if he understands english quite well. The level of difficulty will depend upon his understanding of the other language. There comes a point in one's progress in a foreign language when instead of thinking in the mothertongue and translating it into the new language, we start thinking in the new language itself when dealing with it. That is the level where translation to our primary language becomes almost trivial(barring all the language based limitations ofcourse).
Maybe the future machines will come with some sort of "Mode" for different languages so that they start "thinking" in the language which they are supposed to be translating to the "primary language" just by toggling a couple of switches.
I agree with the rest of your post though.
Re:Great! (Score:3, Interesting)
Actually, this may not be automatic. Japanese and English are (I understand) two languages which differ on fundamental levels. It's not like trying to get to the station and using Franglais: "poor alley a la gar?" It's totally different. Translation is very, very hard, and words are used differently, too. My last job involved working from a spec written in Japanese, and translated by obviously intelligent Japanese people into English. It was... interesting... work, and involved asking the Tokyo office the same question multiple times in different ways, and then cross-referencing the answers, to see if they were consistent.
Also, one of my favourite Japanese poems reads like this:
"She said she would come
At once, and so I waited
Till the moon rose
In the October dawn"
That's from "One Hundred Poems From the Japanese"; I have a different translation in another book, which comes complete with an explanation of over 100 words (I kid you not) of all the different interpretations of the Japanese original, and why that specific translation was selected. It turns out that the original Japanese is written to be gender-ambiguous, and person-ambiguous. It's not "she" and "I", but something like "unresolved consious being" waiting for "mysterious, gender-ambiguous lover-person about whom I care".
You can see why they settled for a snappier version for the poem translation.
So while I love the meloncholy of the translation, I also feel a bit cheated by the fact that my language won't support the original concept, which sounds even better (when you make up words for those concepts and then replay them in slow-mo in your head).
Also, it makes me very glad not to have to translate instruction manuals or business specs.