Intelligent Human-Computer Interfaces
Introduction by
Lynette Hirschman, Guest Editor
This issue is devoted to intelligent human-computer
interfaces -- interfaces that raise the computer to the human
level, rather than requiring that the human adapt to the computer.
An interface may be intelligent because it can communicate
using human language, or because it performs intelligent functions,
or because it adapts to a specific task and user. In all these
cases, an intelligent interface makes interaction with the
computer easier, more intuitive, and more flexible. The area
of intelligent interfaces is just one facet of the broad area
of human-computer interaction. Other important areas include
graphical user interface design, usability, human factors,
visualization, immersive environments, and intelligent tutoring
systems. However, this issue focuses on intelligent interfaces,
representing the spectrum of MITRE activities in this area.
As the computer's role has evolved over the past 50 years,
so have our expectations of the human-computer interface.
Computers were initially used for complex numerical calculations,
such as cryptography or ballistics. Computers were scarce,
large, and expensive. Access was limited to a small set of
highly trained engineers, who understood the computer from
the hardware up.
By the mid-1960s, computers were also routinely used in industry,
for record keeping or transaction processing systems, such
as airline reservations. But computers still cost hundreds
of thousands of dollars and access was restricted to specialists.
No one worried about making computers usable to non-specialists,
anymore than they worried about the human-computer interface
for a slide rule. The computer remained an expensive tool,
used by specially trained technologists.
However, with the 1980s came the "personal computer"
revolution-- the price dropped, availability increased, and
many more people had access to computers. This was followed
in the mid-90s by another revolution -- the World Wide Web.
The Web provided ready access to the Internet and transformed
computers into communications devices, connecting people with
distributed information (e.g., Web-accessible document collections)
and with other people, through e-mail and collaborative computing
environments. Computers had become much smaller, much cheaper,
much more powerful, and much more widely distributed.
Today, computers and computing devices are all around us
-- on everyone's desk at work, on the wrists of supermarket
employees, in the car, and on the telephone. The range of
demands on human-computer interfaces is increasing: preschoolers
play computer games, grade school children are trying out
interactive learning, teens need the Web access to do their
homework, retired people use e-mail to keep in touch with
distant families, and businesses are establishing e-commerce
Web sites. The human-computer interface has suddenly become
a critical element in this revolution. The interface must
make the computer easy to learn, easy to use, and adaptable
to a wide range of users in varied environments.
This short history contains three dimensions that have had
particular importance for human-computer interfaces: the use
by broader, less "expert" segments of society, the
integration of computing into new areas and applications;
and the miniaturization of computers. All of these require
adjustment in the human-computer interface.
One way to make computers accessible to a broader range of
users is to make interaction with a computer more like interaction
with a person. For example, computers might converse using
regular human language, particularly speech. And if computers
were capable of communicating via speech, gesture, or handwriting,
their keyboards could become unnecessary and they could be
made smaller. These changes would allow us a move toward "information
appliances," such as an "intelligent telephone"
that would allow users to "talk" to the Web, as
well as co-workers. Also, such a device would allow us to
retrieve e-mail, receive relevant news feeds, or send a fax.
Future computers might offer the ability to create adaptive
interfaces that can be quickly tailored to specific user needs.
The interface may even be able to learn how to best interact
with a user, just as a person learns how to best interact
with someone they see everyday. Finally, when a computer communicates
via human language and adapts to user needs, users begin to
think of it as an "intelligent agent" or a human
surrogate, instead of thinking of it as a tool. This creates
a new paradigm, where computers are semi-autonomous participants
doing specific tasks for one or more people, such as finding
resources, scheduling meetings, or making travel arrangements.
This issue's articles explore how intelligent interfaces
can close the gap between the user and the computer. Our authors
write about spoken language technology, which allows a person
to converse with a computer; interface flexibility that can
provide output tailored to individual users; and the routine
but useful things "intelligent multimodal participants"
could do in a collaborative human-human environment.
The articles illustrate the range of MITRE technical contributions,
from advanced visionary prototypes, to custom solutions for
sponsor needs, to infrastructure for data collection, and
standards development. All of these articles describe different
ways in which intelligent human-computer interfaces help to
bridge the gap between human and machine -- communicating
in human language, supporting human-human interaction, and
providing flexible user-tailored services.
For more information, please contact guest
editor Lynette Hirschman using the employee directory.
Download a PDF of this issue [1.42MB]