home · mobile · calendar · defenses · 2005-2006 · 

Thesis Defense - Davies

The Efficacy of Personal Knowledge Bases for Materializing Mental Impressions
Stephen Davies
Computer Science PhD Candidate

Software tools abound for managing documents and other information sources, but are rarely used to store the mental knowledge readers glean from reading them. Hence our conceptual understanding -- perhaps our most precious commodity in the so-called "information age" -- is normally left subject to the whims of our erratic memories.

This thesis explores the concept of a personal knowledge base: an experimental database and interface designed to store and retrieve a user's accumulated personal knowledge. It aims to let the user represent information in a way that corresponds more naturally to their mental conceptions than textual notes would. People naturally form mental models of the domains they explore and learn about, and a personal knowledge base allows these to be expressed and archived directly. They need not be converted first to text, a representation which is actually alien to much of the human thought process. A personal knowledge base reflects a user's own subjective understanding of their world, permitting alternate views of the same objective data. And just as each person has many ideas from different domains tied together through perceived associations, it supports an integrated web of knowledge, potentially mirroring the entire contents of its owner's mind.

After defining the notion of a personal knowledge base precisely, and outlining its role in a user's life and expected benefits, this thesis provides the design rationale for a prototype application. It then presents the results of deploying it to twenty volunteers who used it in real-world settings for extended periods of time. The results suggest that such a tool can be an tremendous asset for a variety of knowledge-related tasks, although commitment, discipline, and a willingness to alter one's habits are prerequisites for maximum success. It also sheds some light on the ways people typically work with knowledge, and suggests that unless these patterns are changed, they may ultimately form limitations on the effectiveness of such a tool, and indeed on our use of knowledge in general.

Committee: Roger (Buzz) King, Professor (Chair)
Clayton Lewis, Professor
James Martin, Associate Professor
Kenneth Anderson, Associate Professor
Michael Williams, CaringFamily LLC
Department of Computer Science
University of Colorado Boulder
Boulder, CO 80309-0430 USA
May 5, 2012 (14:20)