Comments on: Recommending BioMedical Articles http://synthese.wordpress.com/2008/03/02/recommending-biomedical-articles/ "Opération qui procède du simple au composé, de l'élément au tout." Mon, 19 Oct 2009 14:46:26 +0000 http://wordpress.com/ hourly 1 By: Daniel Lemire http://synthese.wordpress.com/2008/03/02/recommending-biomedical-articles/#comment-456 Daniel Lemire Wed, 05 Mar 2008 02:54:18 +0000 http://synthese.wordpress.com/?p=82#comment-456 Mathematically, the number of inbound links has to be equal to the number of outbound links. This means that if "4" is the average number of citations, then each paper cites on average 4 papers in your database. I don't know about you, but I typically cite over 15 papers in each one of my papers, so 4 seems very small. Netflix is a bit odd as a data set. 1.2% is pretty dense and that's explained away by the fact that numerous users have rated over a thousand movies. I don't know about you, but this has always seemed odd to me. Mathematically, the number of inbound links has to be equal to the number of outbound links.

This means that if “4″ is the average number of citations, then each paper cites on average 4 papers in your database. I don’t know about you, but I typically cite over 15 papers in each one of my papers, so 4 seems very small.

Netflix is a bit odd as a data set. 1.2% is pretty dense and that’s explained away by the fact that numerous users have rated over a thousand movies. I don’t know about you, but this has always seemed odd to me.

]]>