You are absolutely right pointing to Cyc.
Here is the quote (based on the work Cyc did for Clevelend Clinic) I received a couple days ago in response to my question.
“A quasi-quantifiable example comes from looking at queries done for the Cleveland Clinic (for cohort formation for clinical trials). If one looks at the concepts/assertions touched in order to answer their queries -- most of which focused on finding patients or procedures that met some criteria, rather than abstract questions -- something in excess of 90% of the concepts touched to find the answer were ones that were not added as part of the domain knowledge for that application but were more general background concepts and facts.”
So far, I did not find any study that can better quantify the values, my personal rough estimate for “tribal knowledge” is about 80% and captured knowledge is 20% in a corporate information environment.
I highly appreciate any information on the subject.
Yefim (Jeff) Zhuk
--- On Thu, 5/26/11, doug foxvog <doug@xxxxxxxxxx> wrote:
From: doug foxvog <doug@xxxxxxxxxx>
Subject: Re: [ontolog-forum] "tribal knowledge"
To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
Date: Thursday, May 26, 2011, 3:45 PM
This E-Mail has been scanned for viruses.