http://mendicantbug.com/2008/10/14/relevance-based-language-modeling/ Wanderings into computational linguistics, science, social media and life... Tue, 17 Jan 2012 18:51:53 +0000 hourly 1 http://wordpress.com/ http://mendicantbug.com/2008/10/14/relevance-based-language-modeling/#comment-1072 Fri, 24 Oct 2008 23:40:12 +0000 http://mendicantbug.com/?p=783#comment-1072 I was just thinking about that statement today. Maybe it’s not that old. I guess the 1990 Brown et al paper on statistical MT is a classic. Maybe the fact that I am only reaching back 18 years is telling, since Sapir & Co. was much earlier. We do frequently go back to the Shannon paper on information theory, but that’s not really nlp.

]]> http://mendicantbug.com/2008/10/14/relevance-based-language-modeling/#comment-1070 Fri, 24 Oct 2008 23:28:18 +0000 http://mendicantbug.com/?p=783#comment-1070 If 2001 is old, maybe NLP has jumped the academic shark, hehe. There is somethign to be said for classics. Linguists still read early Sapir, Fillmore, Dowty. Is there a classic work in NLP that never goes out of date?

]]>