Position Paper for CHI 96 Basic Research Symposium (April 13-14, 1996, Vancouver, BC)


Professor Raoul N. Smith
College of Computer Science
Northeastern University, 219CN
Boston, MA 02115


I have become interested in the practical use of information about emotions in HCI. My previous experience includes extensive research in natural language processing at Northwestern University and in intelligent interfaces while at GTE Laboratories. I was Co-Chair of SIGCHI (1982-85) and Chair of CHI83 and Co-Chair of CHI86.

My interest in emotions developed through my own psychotherapy and the subsequent development of two computerized aids to psychiatry:

Dr. Bob
a system that retrieves stories from the Big Book of Alcoholics Anonymous and books on childhood sexual abuse. It is a memory- based reasoning system that runs on the parallel processor Thinking Machines, Inc. CM2.
a system to help people suffering from alexithymia, an affective disorder in which the patient is unable to articulate his/her feelings.

I have been interested in the formal representation of emotions and have developed a frame-like data structure (used in Alex). I am currently building a large database of emotions for applied systems that can put them to use.

I would like to participate in the Basic Research Symposium by presenting my data structure for emotions and then asking the participants to brainstorm about the structure and systems. This would aid in directing future research.

Obvious uses of emotions in computer systems include

The uses in interface design and in computer suported cooperative work are not that obvious. C. Rich (SIGCHI Bulletin, January 1966), for example, uses anthropomorphism (which I would extend to emotions!) as a guide in his development of software interface agents. Can emotions play a role in such agents and what would that role be? What emotions are appropriate in HCI? When is the knowledge of emotions appropriate to HCI?...