Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!samsung!zaphod.mps.ohio-state.edu!uwm.edu!dogie.macc.wisc.edu!decwrl!shlump.nac.dec.com!baviki.enet.dec.com!good From: good@baviki.enet.dec.com Newsgroups: comp.cog-eng Subject: Re: reference wanted (or refs.) Message-ID: <8635@shlump.nac.dec.com> Date: 23 Feb 90 16:50:23 GMT References: <3225@servax0.essex.ac.uk> Sender: newsdaemon@shlump.nac.dec.com Reply-To: good@baviki.enet.dec.com () Organization: Digital Equipment Corporation Lines: 14 There's been lots of work done on this topic. The research that I find most interesting, and that also seems relevant to your question, is the statistical semantics work done at Bellcore. One older reference is the article on "Statistical Semantics: Analysis of the Potential Performance of Key-word Information Systems" by George W. Furnas, Thomas K. Landauer, Louis M. Gomez, and Susan T. Dumais. It's published in the Bell Systems Technical Journal, Vol. 66, No. 6, Part 3, pp. 1753-1806, 1983. I know they have more recent articles, but I don't have references to them offhand. Do a search for articles by the 4 authors above and you should find some of that work. Michael Good Good@Baviki.Enet.Dec.Com