Comments on: Introduction to Statistical Machine Translation http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/ Tue, 09 Mar 2010 22:14:18 +0000 http://wordpress.org/?v=2.9.1 hourly 1 By: axiomatic | The Nation of Automated Changing Translation: Fearful http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-23559 axiomatic | The Nation of Automated Changing Translation: Fearful Sun, 21 Jun 2009 03:53:27 +0000 http://michaelnielsen.org/blog/?p=577#comment-23559 [...] Translate is a statistical translation system, based on analyzing parallel documents that have been professionally translated by people. For [...] [...] Translate is a statistical translation system, based on analyzing parallel documents that have been professionally translated by people. For [...]

]]> By: links for 2009-03-27 « My Weblog http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-19159 links for 2009-03-27 « My Weblog Sat, 28 Mar 2009 04:03:41 +0000 http://michaelnielsen.org/blog/?p=577#comment-19159 [...] Michael Nielsen » Introduction to Statistical Machine Translation (tags: ai nlp) [...] [...] Michael Nielsen » Introduction to Statistical Machine Translation (tags: ai nlp) [...]

]]> By: Science Blogging Benefits Everyone at Synthesis http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-19041 Science Blogging Benefits Everyone at Synthesis Wed, 25 Mar 2009 07:12:58 +0000 http://michaelnielsen.org/blog/?p=577#comment-19041 [...] that mainstream journalists either can’t or wont fill - that of debunking pseudoscience. Michael Nielsen’s post on Statistical Machine Translation was written as he’s learning, making it a [...] [...] that mainstream journalists either can’t or wont fill – that of debunking pseudoscience. Michael Nielsen’s post on Statistical Machine Translation was written as he’s learning, making it a [...]

]]> By: Michael Nielsen » Implementing Statistical Machine Translation Using MapReduce http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-18999 Michael Nielsen » Implementing Statistical Machine Translation Using MapReduce Tue, 24 Mar 2009 01:23:03 +0000 http://michaelnielsen.org/blog/?p=577#comment-18999 [...] post is a followup to an earlier post about the beautiful subject of statistical machine translation. In the earlier post I described how to build a statistical model that can be used to automatically [...] [...] post is a followup to an earlier post about the beautiful subject of statistical machine translation. In the earlier post I described how to build a statistical model that can be used to automatically [...]

]]> By: Michael Nielsen http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-18614 Michael Nielsen Fri, 13 Mar 2009 14:54:51 +0000 http://michaelnielsen.org/blog/?p=577#comment-18614 Gil - Yes, that's right. More recent iterations of statistical machine translation incorporate much more linguistics. But it seems like you get a much bigger gain by starting with statistical ideas, and then later grafting on the linguistic ideas, not the other way round. I found this very surprising. Gil – Yes, that’s right. More recent iterations of statistical machine translation incorporate much more linguistics. But it seems like you get a much bigger gain by starting with statistical ideas, and then later grafting on the linguistic ideas, not the other way round. I found this very surprising.

]]> By: Gil Kalai http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-18613 Gil Kalai Fri, 13 Mar 2009 14:48:20 +0000 http://michaelnielsen.org/blog/?p=577#comment-18613 Dear Michael, is it the case that this method involves no parsing? (And very little linguistics.) Dear Michael, is it the case that this method involves no parsing? (And very little linguistics.)

]]> By: Michael Nielsen http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-18597 Michael Nielsen Thu, 12 Mar 2009 19:30:34 +0000 http://michaelnielsen.org/blog/?p=577#comment-18597 Rajiv Das - No, I'd never heard of the book before; my post was based on the original papers. But based on Peter Norvig's review, I'll take a look at the book. Rajiv Das – No, I’d never heard of the book before; my post was based on the original papers. But based on Peter Norvig’s review, I’ll take a look at the book.

]]> By: Michael Nielsen http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-18596 Michael Nielsen Thu, 12 Mar 2009 19:28:37 +0000 http://michaelnielsen.org/blog/?p=577#comment-18596 Or - Great catch. I'll fix it. Or – Great catch. I’ll fix it.

]]> By: Rajiv Das http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-18587 Rajiv Das Thu, 12 Mar 2009 03:08:13 +0000 http://michaelnielsen.org/blog/?p=577#comment-18587 Are you reading the last chapter of this book http://www.amazon.com/Language-Processing-Prentice-Artificial-Intelligence/dp/0131873210/ref=pd_bxgy_b_text_b Good One. Are you reading the last chapter of this book

http://www.amazon.com/Language-Processing-Prentice-Artificial-Intelligence/dp/0131873210/ref=pd_bxgy_b_text_b

Good One.

]]> By: Or http://michaelnielsen.org/blog/introduction-to-statistical-machine-translation/comment-page-1/#comment-18585 Or Thu, 12 Mar 2009 02:49:13 +0000 http://michaelnielsen.org/blog/?p=577#comment-18585 Great post (as usual) Regarding estimating Pr(e_2|e_1), didn't you forget to divide by an estimator for Pr(e_1)? Taking #(e_1,e_2)/N gives you an estimate of the joint prob. Pr(e_1,e_2) Great post (as usual)

Regarding estimating Pr(e_2|e_1), didn’t you forget
to divide by an estimator for Pr(e_1)?
Taking #(e_1,e_2)/N gives you an estimate of the joint prob. Pr(e_1,e_2)

]]>