tag:blogger.com,1999:blog-6569681.post5849894149422340842..comments2024-01-15T13:17:33.771-08:00Comments on Geeking with Greg: Effectiveness of personalized searchGreg Lindenhttp://www.blogger.com/profile/09216403000599463072noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-6569681.post-25714092205710430992008-02-03T07:59:00.000-08:002008-02-03T07:59:00.000-08:00Thanks for the links to these papers. This papaer ...Thanks for the links to these papers. This papaer <A HREF="http://portal.acm.org/citation.cfm?id=1331740.1331793&coll=&dl=GUIDE&type=series&idx=SERIES11154&part=series&WantType=Proceedings&title=WI" REL="nofollow">Experimental Bounds on the Usefulness of Personalized and Topic-Sensitive PageRank</A> from Web Intelligence 2007 has some interesting results too.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6569681.post-68567969369094790592007-08-10T17:38:00.000-07:002007-08-10T17:38:00.000-07:00I pick on Google's personalized search for primari...<I>I pick on Google's personalized search for primarily being a profile-based technique focused on long-term history ([1] [2] [3]) and argue for a fine-grained, click-based approach.</I><BR/><BR/>And to beat my usual drum, the ultimate short term, fine-grained, click-based approach is the 30-year-old notion of relevance feedback. <BR/><BR/>It is really surprising to me that it has taken us so long to come back to where we began, where we should have been all along.<BR/><BR/>I think one of the first things I ever wrote on your blog, Greg, was a pointer to a paper from 2003 where an old colleague of mine had done profile-based personalization and found that it didn't work.<BR/><BR/>I'm not as good at keeping track of my old posts as you are, though. I'm guessing this is somewhere around April of 2006 that I made that comment?<BR/><BR/>Anyway, I'm not surprised.Anonymousnoreply@blogger.com