[983] in Public-Access_Computer_Systems_Forum
RE: Boolean & Probabilistic
daemon@ATHENA.MIT.EDU (KINGH%SNYSYRV1.BITNET@ricevm1.rice)
Thu Aug 13 10:47:45 1992
Date: Thu, 13 Aug 1992 09:41:59 CDT
Reply-To: Public-Access Computer Systems Forum <PACS-L%UHUPVM1.BITNET@ricevm1.rice.edu>
From: KINGH%SNYSYRV1.BITNET@ricevm1.rice.edu
To: Multiple recipients of list PACS-L <PACS-L%UHUPVM1.BITNET@ricevm1.rice.edu>
----------------------------Original message----------------------------
Who determines relevance? The definition of relevance in ranked retrieval
systems is the number of times the keyword occurs, right? Is there a newer,
more sophisticated, more valid and reliable method of assessing relevance?
Also, has anyone validated methods that are supposed to predict relevance? And,
are these methods reliable, i.e., over time, over users, they consistently
and correctly predict which records will be relevant for users?
Hannah King
SUNY HSC Library Syracuse
kingh@snysyrv1
kingh@vax.cs.hscsyr.edu
