Comments on: Costs and Transparency in Ranking Systems http://wanderingstan.com/2007-03-12/costs_and_transparency_in_ranking_systems wanderings and wonderings of stan james Tue, 16 Mar 2010 00:42:30 +0000 http://wordpress.org/?v=2.7.1 hourly 1 By: Joe http://wanderingstan.com/2007-03-12/costs_and_transparency_in_ranking_systems/comment-page-1#comment-72 Joe Tue, 30 Nov 1999 00:00:00 +0000 #comment-72 Only global ratings suffer from it. Subjective metrics are a lot more attack resistant - http://www.sigcomm.org/sigcomm2005/paper-CheFri.pdf Only global ratings suffer from it. Subjective metrics are a lot more attack resistant - http://www.sigcomm.org/sigcomm2005/paper-CheFri.pdf

]]> By: stan http://wanderingstan.com/2007-03-12/costs_and_transparency_in_ranking_systems/comment-page-1#comment-73 stan Tue, 30 Nov 1999 00:00:00 +0000 #comment-73 Yes! The problem is when you want a global "objective" ranking system. When you want ratings with a known bias (e.g. the academy awards, your friend's delicious tags) then these problems go away. I guess I should do a follow up post about how Lijit does search in a subjective way, thus avoiding the google problems. Thanks for the paper reference, Joe...I'll read it tonight. Yes! The problem is when you want a global “objective” ranking system. When you want ratings with a known bias (e.g. the academy awards, your friend’s delicious tags) then these problems go away.

I guess I should do a follow up post about how Lijit does search in a subjective way, thus avoiding the google problems.

Thanks for the paper reference, Joe…I’ll read it tonight.

]]>