This paper is a good illustration of what I see as a critical shortcoming of modern NLP work: it is all about math and algorithms, there is nothing about actual language.<p>The paper has lots of information about: neural network architectures, parameter update equations, learning rules, and inference algorithms.<p>There is nothing about: part of speech categories, relative clauses, morphology, affixes or compound words, the theta criterion, the content/function distinction, verb tenses, agreement, or anything else related to actual linguistic phenomena.<p>To me this seems a bit like the Aristotelian thinkers who tried to reason about physics based on pure mathematical analysis, without any empirical work.
I hate when people judge machine translation progress by looking at Google translate. Google translate is really old and is no longer state of the art. But the best neural network based systems like this are much too expensive to use in production, at least for free. They say their new neural network ASICs will make it more practical, at least.<p>Anyway the scale of these neural nets is quite incredible. Google is getting far ahead of what any individual researcher with a few consumer GPUs can do.
Related blog post: <a href="https://research.googleblog.com/2016/09/a-neural-network-for-machine.html" rel="nofollow">https://research.googleblog.com/2016/09/a-neural-network-for...</a>
60% better than Google translate still is at best elementary school level fluency in a language. It’s a needed improvement, but still far from good.<p>It’s crazy how bad Google translate is, try it with any German text, and you’ll get 90% ununderstandable garbage out.<p>EDIT: Seriously, downvotes for this?<p>Try this: <a href="https://translate.google.com/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http://www.spiegel.de/politik/ausland/bratislava-eu-verteidigungsminister-wollen-enger-zusammenarbeiten-a-1114151.html" rel="nofollow">https://translate.google.com/translate?sl=auto&tl=en&js=y&pr...</a><p>What’s " If you look at Ursula von der Leyen and Jean-Yves Le Drian before the meeting of defense ministers in Bratislava, one might think that Berlin and Paris would never prefer liked." supposed to mean?<p>Or "Individual states can not prevent the European Council may decide by a qualified majority, the SSZ."
"In addition to releasing this research paper today, we are announcing the launch of GNMT in production on a notoriously difficult language pair: Chinese to English.... we will be working to roll out GNMT to many more of these over the coming months."<p>interesting news for all the translation startups... like yc's unbabel.
Dumb question: Why doesn't Google Translate hardcode perfect translations for 1,000,000 of the most popular requests?<p>So many times it screws up really basic, common phrases and I'm always wondering how is that bar that low?
"Using a human side-by-side evaluation on a set of isolated simple sentences, it reduces translation errors by an average of 60% compared to Google's phrase-based production system."<p>Does this mean that after checking the translation by a human it becomes 60% better than the phrase-based production system? (by which they presumably mean Google Translate...?). That seems rather disappointing.
I've been attempting to use this system for a while today to converse with a native Chinese speaker and it seems to be better but still very far from human-level translation. Maybe they've been using especially bad human translators in their comparisons?