IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS, VOL. SMC-12, NO. 4, JULY/AUGUST 1982
[11] A. deGroot, "Perception and memory versus thought: Some old
ideas and recent findings," in B. Kleinmuntz, Ed., Problem Solving.
New York: Wiley, 1966.
[12] David K. Dirlan, "Most efficient chunk sizes," Cognitive Psychol.,
vol. 3, 1972.
[13] J. J. Donovan, "Database system approach to management decision
support," ACM Trans. Decision Syst., vol. 1, no. 4, Dec. 1976.
[14] P. W. Frey, Ed., Chess Skill in Man and Machine. New York:
Springer-Verlag, 1977.
[15] P. W. Frey and P. Ademan, "Recall memory for visually presented
chess positions," Memory and Cognition, vol. 4, no. 5, 1976.
[16] E. F. Lindquist, Design and Analysis of Experiments in Psychology
and Education. Boston: Houghton Mifflin, pp. 266-273, 1953.
[17] M. Newborn, Computer Chess. New York: Academic, 1975.
[18] A. Newell and H. A. Simon, Human Problem Solving. Englewood
Cliffs, NJ: Prentice-Hall, 1972.
[19] J. S. Reitman, "Skilled perception in GO: Deducing memory struc-
tures from inter-response times," Cognitive Psychol., vol. 8, no. 3,
pp. 336-356, July 1976.
[20] F. Restle and E. Brown, "Organization of serial pattern learning,"
in G. H. Bower, Ed., The Psychology of Learning and Motivation.
vol. 4, New York: Academic, 1970.
[21] A. L. Samuel, "Some studies in machine learning using the game of
checkers II, Recent progress," IBM J. Res. Develop., vol. 11, no. 6,
1967.
[22] V. Slamecka, H. Camp, A. N. Badre, and D. Hall, "MARIS: A
knowledge system for internal medicine," J. Information Manage-
ment and Processing, vol. 13, pp. 273-276, 1977.
[23] A. L. Zobrist and F. R. Carlson, Jr., "An advice-taking chess
computer," Sci. Amer., vol. 228, pp. 92-105, 1973.
A Trainable On-Line Model of the Human
Operator in Information
Acquisition Tasks
AZAD M. MADNI, MEMBER, IEEE, MICHAEL G. SAMET, AND AMOS FREEDY, MEMBER, IEEE
Abstract-A trainable model of the human operator in information
acquisition tasks is described. The purpose of this model, called the
adaptive information selector (AIS), is to select and present textual mes-
sages automatically to users in computer-based tactical systems. AIS
algorithms, which control and present messages in order of priority in real
time, are based on multiattribute characterization of tactical messages. The
AIS employs an adaptive pattern recognition model to "learn" user prefer-
ence structure incrementally during actual task performance. Across each
of the command situations, the priority of messages is determined by the
AIS in accord with the information selection behavior exhibited by the user
in the model-training mode. The AIS was implemented and tested in a
simulated environment. The results demonstrated the model's capability to
1) converge on distinctive information processing strategies exhibited by
different operators; and 2) effectively present messages in their order of
priority for each operator. The AIS is potentially useful in performing
information distribution functions in command and control systems and in
aiding the performance of personalized searches of large data bases.
I. INTRODUCTION
T ECHNOLOGICAL advances have led to rapid in-
creases in the amount and transfer rate of military
information. Today information must be processed more
efficiently and more effectively for commanders to make
tactical decisions responsive to the rapidly changing
succession of events. To meet this need, new computer-
Manuscript received November 1, 1981; revised March 5, 1982. This
work was supported by Defense Advanced Project Agency Cybernetics
Technology Office under Contract MDA903-78-C-0127.
The authors are with Perceptronics, Inc., 6271 Variel Avenue, Wood-
land Hills, CA 91364.
based systems for command, control, and communications
(C3) are being continually developed, implemented, and
evaluated. These systems are intended primarily to aid in
the collection, processing, and utilization of different types
and amounts of military data. The overall process is cyclic
-as information is being used, other information is being
processed, and new information is being sought and col-
lected. In this environment the dynamics of information
flow is of critical importance and must be constantly
monitored and directed.
The consensus on current computer-based military sys-
tems for C3 operations is that the message traffic within
the network has increased to such an extent as to over-
whelm a commander and his staff. New techniques to
control information flow are required to best match system
capability with human operator characteristics. Review of
previous research has suggested that a significant step in
this direction would be to individualize and automate
information selection [1]. The goal is to allow each user to
obtain information consistently that is both relevant and
timely with regard to his individual processing characteris-
tics and immediate decisionmaking needs. Considering the
large number of users in a typical C3 system, the effect of
individualized message handling systems on total system
performance would be to both increase throughput and
improve decisionmaking performance.
Thus the purpose of the research undertaken here was to
demonstrate means by which computer-based models of
0018-9472/82/0700-0504$00.75 (©1982 IEEE
504