DYU@NCCIBM1.BITNET (06/14/89)
The Heuristic, newsletter of the IEEE Task Force on Expert Systems
vol. 1, issue 1, Summer 1989
BITNET DYU@NCCIBM1
[This is the full text for the first newsletter of the IEEE
Computer Society Task Force on Expert Systems]
----- MASTHEAD -----
About our Name...
We have chosen THE HEURISTIC as the title of the Task Force on Expert
Systems Newsletter. Not only does this reflect that expert systems
are tools for applying heuristics, but also that the Task Force is
an organization for discussing heuristics about expert systems in
practice.
Editor: Rodger Knaus, (202) 966-2582
Research and reporting: Sandra Hoffman, (202) 226-6849
Layout and Design Joe Hooper, Lori Rottenberg
Distribution Judy Lamont
Deadline for the Fall '89 issue: Aug. 1, 1989
Send newsletter contributions to: Rodger Knaus, Instant Recall,
5900 Walton Rd., Bethesda, Md. 20817.
Preferred format: a machine-readable file, either an ASCII file
on a 5.25 IBM-PC/AT diskette, or an upload on the
Instant Recall BBS. (301)-983-8439
BBS info: Up during business hours and some other times; runs
under RBBS at 300 to 2400 baud, with 8 data bits, 1
parity bit and null parity. This BBS is for file
transfer, NOT for general messages.
----- NEXT MEETING -----
The next meeting of the IEEE Task Force on Expert Systems
will be July 25, 9-11 a.m., at the American Management Society
(AMS). After a 20-min. business meeting, several speakers will
present expert systems applications in their organization. AMS is
located in Roslyn, VA (in the Washington, D.C. metropolitan area)
at 1777 N. Kent St., 2 blocks from the Roslyn Metro stop.
----- EDITORIAL -----
Welcome to the IEEE Task Force on Expert Systems.
Maybe your reaction was like mine: What is a Task Force on
Expert Systems anyway, and who needs one? There are already too
many conferences and publications on compters, and expert systems
in particular. Why have more?
I saw one possible answer at "On the Art of Fixing a Shadow: 150
Years of Photography." Since photography is 5 times older than
expert systems, its path as a technology suggests how expert
systems may evolve.
The first photographers built their own cameras, made their own
film, and printed their own pictures. Camera and film science
dominated the early years of photography. However, as the basic
technology matured, craft displaced science as the prime concern
of photographers. Today, photographers use cameras, film and
expertise to create images for people. In a similar transition,
knowledge engineers are emerging as craftspersons who use
hardware, software and skill to create decision tools for people.
Photography could not exist without lens design and film
chemistry; architecture could not exist without structural
engineering; and knowledge engineering could not exist without
successful and continuing expert systems research. However, like
architecture and photography before it, knowledge engineering
today is a design rather than a research activity.
The concerns of the knowledge engineer are those of a designer
rather than a scientist. The knowledge engineer is an architect
of software, with the architect's concern for building something
on time and within budget, meeting the current and future needs
of the client, and creating an esthetically pleasing environment
where form reflects function.
This is where the Task Force comes in. The Task Force is a forum
for knowledge engineers to
* see and present current work
* discuss new tools and techniques
* learn how things worked in practice.
Our focus is on how things work in the field rather than in the
lab; on the lifecycle rather than the idea and prototyping stage;
on the complete system rather than the isolated design attribute;
and on design and architecture rather than basic science and
engineering.
The Task Force plans meetings and conferences about expert
systems in practice. We plan standards that may eventually help
expert systems communicate. The nature and direction of the Task
Force depends on you, the working knowledge engineer.
Our newsletter also will become what you make it. Do you have a
more graceful name for the organization than IEEE Task Force on
Expert Systems? I'll publish a list of the best suggestions in
the next newsletter. Did an expert system application turn out
unexpectedly good or bad, easy or hard? Write it up for our "How
Did it Work" column. Was there an event of interest in
your area? Share it with the Task Force nationally in a
newsletter article.
This is your organization. We invite and welcome your
participation.
----- MESSAGE FROM THE CHAIR -----
by: Dan Yurman
202-475-6754
BITNET DYU@NCCIBM1
Welcome to the Task Force on Expert Systems
The IEEE Computer Society announces the formation of a Task Force
on Expert Systems Applications. The purpose of the Task Force is
to support interest in the development and use of expert systems
applications. The IEEE Computer Society, which has over 100,000
members worldwide, authorized the creation of the Task Force at a
meeting of its Technical Activities Board (TAB) held March 2nd in
San Francisco, CA.
The objectives of the Task Force are to improve the abilities of
organizations and individuals to work with expert systems
technologies. The Task Force will sponsor activities at the
national and international levels, but also support local events.
The Task Force will publish a newsletter, exchange electronic
mail, provide speakers for conferences, organize tutorials,
symposiums, and convene meetings on standards either on its own
initiative or jointly with other IEEE Computer Society functions.
The Importance of Expert Systems
MIS and end users are experiencing an explosion of interest and
activity in expert systems applications in almost all sectors of
business, government, and education. This is taking place due to
the wide distribution of expert system shells on personal
computers and workstations. It is a remarkable change in the
field of artificial intelligence in which developers usually rely
on specialized computer platforms and programming languages.
Expert systems are the most mature and resilient products to
emerge from the AI community, and they are being adopted by
corporations and government departments to improve productivity.
They are doing this because the applications of expert systems to
specific knowledge intensive systems return high yields. Success
stories for expert systems are more common now than two years
ago. A current estimate is there are 2,000 operational expert
systems and 80% of them on running on personal computers.
The Value of the Task Force
The greatest value which will be derived from participating in
the Task Force will come from regular discussions among Task
Force participants. In some ways, this will resemble the
informal interactions of a user group. In other ways it will
compliment many of the professional activities of the IEEE
Computer Society.
Within these broad themes, there are many diverse interests,
including business activities such as banking and finance,
manufacturing and service functions, medical practice, government
functions, including the military, and education. These interests
will be addressed through locally sponsored and nationally
significant activities including conferences, workshops,
lectures, a newsletter, and other appropriate mechanisms.
Membership
All meetings of the Task Force are be open to the public and will
be announced ahead of time in the news media. Anyone who has an
interest in the objectives of the Task Force is invited to attend
its functions and participate in its activities. Membership in
the IEEE Computer Society is not required to attend our meetings.
Since the Task Force is oriented toward development and use of
expert systems applications, we expect and encourage the interest
of vendors of computer hardware, software, and services. Future
activities of the Task Force will be developed consistent with
the goals and objectives of the IEEE Computer Society.
First Meeting
On April 21, 1989, the Task Force held it's kickoff meeting in
Washington, DC. A total of 61 people attended the session. The
meeting was held at the Departmental Auditorium of the Internal
Revenue Service through the good offices of Ted Rogers, Director
of the IRS Artificial Intelligence Center. The Task Force
created and staffed five Vice-Chair positions: Standards,
Conferences, Communications, Industry Relations, and Local
Meetings.
We have two vacant vice chair positions - treasurer and
membership. Here is what they do.
Treasurer: Responsible for keeping track of Task Force finances,
for fund raising including setting up a means to cover costs, and
solicitations of "in kind" contributions.
Membership: Responsible for developing and implementing a
membership recruitment campaign for the Task Force working at the
national level. Works with other Vice-Chairs to carry out these
functions.
More than 400 people have expressed an interest in the Task
Force. These names have been submitted to the TAB Coordinator
for creating of a mailing list. Publicity about the Task Force
has appeared in several computer trade publications and
approximately 60 people called for more information as a result.
Texas Instruments Corp. provided a mailing list of people who
attended their AI Satellite Symposium last November, and the
company is planning to provide publicity about the Task Force in
an upcoming newsletter. Press releases were sent to IEEE Expert
and IEEE Computer Magazine.
Persons desiring to be placed on the mailing list should send
their name and address to:
Task Force on Expert Systems
c/o TAB Coordinator
IEEE Computer Society
1730 Massachusetts Ave. NW
Washington, DC 20036
----- CONTACTS -----
IEEE Task Force Contacts
Who to contact:
To be put on the Expert Systems Task Force mailing list:
TAB Coordinator
IEEE Computer Society
1730 Mass Ave. NW
Washington, DC 20036
For Task Force sponsorship of local activities:
Chair, IEEE Expert Systems Task Force
Other: The appropriate chair, from the list below.
Chair: Dan Yurman
Task Force on Expert Systems
c/o TAB Coordinator
IEEE Computer Society
1730 Mass Ave. NW
Washington, DC 20036
(202) 475-6754 (9-5 EST)
MCI Mail: 364:1277
IEEE C.S. Comp-Mail: CMP5845
BITNET DYU@NCCIBM1
Vice Chair for Communications
Sandra Hoffman
Congressional Budget Office
2nd & D, SW, Rm. 450
Washington, D.C. 20515
(202) 226-2775
Vice Chair for Conferences
Jerry Feinstein
Phase-Linear Systems/ICF
9300 Lee Highway
Fairfax, VA 22031
(703) 934-3280
Vice Chair for Industry Relations
Joseph Schmuller
CDM Federal Programs Corp.
13135 Lee Jackson Highway
Fairfax, VA 22033
(703) 968-0900
Vice Chair for Meetings
Randy Manner
American Management Systems
1777 N. Kent St.
Arlington, VA 22209
(703) 841-6849
Vice Chair for Standards
Capt. Dave Howell
Artificial Intelligence
Program Management Office
U.S. Air Force Logistics Command
HQ AFLC/MM-AI
Wright Patterson AFB
Ohio 45433
(513) 257-2571
HOWELL@WPDIS01.ARPA
Vice Chair for Membership: vacant - contact the TAB coordinator
Treasurer: vacant - contact the Chair
Newsletter Editor
Rodger Knaus
Instant Recall
5900 Walton Rd.
Bethesda, Md. 20817
(301) 530-0898
BBS: (301) 983-8439
----- LOCAL -----
Local events and meetings are one of the most important
ways in which expert systems professionals can exchange
information and grow professionally. The IEEE Task Force on
Expert Systems will sponsor local events about expert systems
applications, provided that these events are appropriate to a
professioal organization under IEEE's guidelines; what this means
in practice is that sponsored events are of interest to expert
systems professionals and are not vehicles for sales and
marketing activities. To obtain Task Force sponsorship of an
event, contact the Chair, Dan Yurman, at (202) 475-6754.
----- INDUSTRY -----
Joseph Schmuller
Expert Systems Team
CDM Federal Programs Corporation
Fairfax, VA
Our objective, broadly stated, is to liaison with industry -- to
get industry involved with the activities of the Task Force. We
started by analyzing the world that we will liaison with, and
distinguished first between "Industrial Expert Systems" and the
"Expert Systems Industry" -- in other words, between consumers of
Expert Systems technology and vendors of the technology.
Both consumers and producers can be represented by a 2 X 2
matrix. Consumers may be either public or private sector, and
either "system developers" or "system users" (we recognize that
these two roles are not necessarily mutually exclusive). Our
"Consumers" matrix, then, encompasses "Sector" (Public and
Private) and "Role" (Developer and User). As we contact members
of the Consumer community, two questions will be answered: How
is the technology being used? How can the Task Force help users
of the technology?
Now, let's look at the "Vendors" matrix. Manufacturers of Expert
System Tools necessarily distinguish between applications that
reside on Personal Computers and those that reside on larger
platforms, and they design their tools accordingly. Our Group
will establish connections with vendors from both arenas.
Another distinction is between hardware tools and software tools.
Thus, our "Vendors" matrix is defined by "Platform" (PC and
non-PC) and "Type of Tool" (Hardware and Software). Our contacts
with the Vendor Community will result in reviews of new products.
We are also interested in influencing the development of future
products.
Committees
In order to carry out Group objectives, we'll have four
committees, and these committees reflect the aforementioned
analysis. I've selected three committee chairpersons (the fourth
is yet to be named), and I'll work with them to find members for
each committee. The committee members will initiate contacts
with the Consumer and Vendor Communities, and the appropriate
chairperson will co-ordinate these contacts.
For the Consumer Community, Don Ramsey (of the law firm of
Sutherland, Asbill, and Brennan) will chair the Private Sector
Committee, and Bruce Ramsay (of the IRS) will chair the Public
Sector Committee. I'll work with Don and Bruce to find at least
four members for each committee. Each committee member will
contact either users of Expert Systems or developers of Expert
Systems. The purpose of the contacts is to find out the types of
systems being developed, who is doing the development, and how
Expert Systems are being used within a particular organization.
For the Vendor Community, Rod Pitts (of the AI Special Interest
Group of the Capital Area PC Users Group) will chair the
PC-based Tools Committee. At present, the chairperson of the
Non-PC Tools Committee is unnamed. I'll work with Rod and the
fourth chairperson to find at least four members for each
committee. Each committee member will contact vendors of
hardware or vendors of software. The contacts will be for the
purpose of finding out about new releases and for receiving
review copies of products. Each committee member will review the
products that he or she receives.
My thanks to Don, Bruce, and Rod!
Meetings
I plan to hold chairpersons' meetings in odd-numbered months and
Industry Relations Group Meetings in even-numbered months. At
each chairpersons' meeting, we will identify a speaker or set of
speakers for the subsequent Industry Relations Group meeting.
The speakers will be drawn from the Consumer and Vendor
Communities.
At our chairpersons' meetings, we will also begin to organize
poster sessions, presentations, and other events for meetings of
the entire Task Force. To facilitate the production of these
events, the Vice Chairman and the committee chairpersons will
ask for the help of Group members who are not members of any of
the four committees.
----- CONFERENCE -----
FAST START ON CONFERENCES PLANNING
by Jerry Feinstein,
Chair, Conference Working Group
ICF/Phase Linear
tel. 703-934-3280
fax. 703-934-9740
The Conference Working Group met on 4/21/89 to develop a plan for
identifying and implementing its activities. Our long-range goal
is to establish a yearly conference that would focus on expert
systems with an emphasis on the user and manager. A more
immediate goal is to establish informal workshops where managers
and system builders could discuss lessons learned, successful
approaches used elsewhere, etc.
The Conference Group identified a number of 'interest areas'
around which might spring ideas for workshops, tutorials, and
tracks for future conferences: knowledge acquisition;
validation, verificaiton, and testing; project management and
implementation of E.S.; business applications of E.S; PC-based
E.S.; statistical applications; and standards. These topical
groupings are open to change and expansion, guided by the
interests of the Task Force at large. People are free work on as
many groups as their time and interests permit.
For the future, the Conference Working Group selected MANAGING AN
EXPERT SYSTEM PROJECT/PROGRAM as a general theme that could
develop into workshops or conferences.ed From this theme, tracks,
tutorials, and topics would flow logically and almost
sequentially.
From the theme, we developed a preliminary structure based on
input from our group. We call this structure preliminary because
its purpose is to provide a point of departure or skeleton on
which others in our Task Force may build.
Staffing an E.S. Group
Selecting the Right Application
Selecting Appropriate Shells
Cost-Justifying E.S.s
Managing the Development
Institutionalization
Finding a Project Champion
Developing User Groups
Validation
Measuring Productivity
Maintenance
----- COMMUNICATIONS -----
NEWSLETTER STARTED, BBS PLANNED
The Communications Committee has been busy with the first issue
of a planned quarterly newsletter. Publication of each issue
will be around the start of a season (e.g. June 15 for the first
issue).
Under the guidance of our chair, Sandra Hoffman, the newsletter
has been a pipelined effort, with reporting by Sandra Hoffman,
editing by Rodger Knaus, design and layout by Joe Hooper and
distribution by Judy Lamont, with the help of the IEEE Technical
Activities Board Coordinator, Lori Rattenberg.
Research and reporting: Sandra Hoffman, (202) 226-6849
Editing: Rodger Knaus, (202) 966-2582
Layout and Design Joe Hooper
Publicity Judy Lamont
Many members of the Task Force have expressed interest in an
electronic bulletin board. Now that the first newsletter is out,
the Communications Committee wants to start planning the BBS. We
need ideas and help in every aspect of the bulletin board,
including donations of hardware and software, bulletin board
expertise, and ideas about structure and content of the board; in
return for your contributions, you will be able to post some
information about your professional services on the board. If
you want to contribute to the BBS, leave a message to Rodger
Knaus (SYSOP) on the Instant Recall BBS, (301) 983-8439. the
board is up during business hours, plus some evenings and
weekends, operates under RBBS at 300 to 2400 baud, with 8 data
bits, 1 parity bit and null parity.
----- STANDARDS -----
KBS LIFECYCLE MAY BE STANDARDS FOCUS
by Capt. Dave Howell, Standards Vice-Chair
HOWELL@WPDIS01.ARPA
513-257-2925
The turnout for the Standards Committee portion of the first IEEE
Task Force on Expert Systems meeting was excellent. Individuals
from 15 organizations attended; they came from both the public
and private sector, and included both end users and
knowledge-based system (KBS) manufacturers.
In spite of this diversity, when the attendees said why they were
interested in standards, widespread interest on the KBS life
cycle emerged. The group saw standards as a way to reduce life
cycle costs, because with standards you can
* develop good KBS costing models
* provide effective development methodologies
* determine proper verification and validation (V&V) techniques
The group also wanted "lessons learned" to be part of the
standards process.
A consensus developed that the scope of the Standards Committee
should be at a high enough level to allow for language
independence. The group did not see itself getting initially
involved in low level implementation issues, i.e. Prolog versus
Lisp versus shells, etc. Further, the high-level focus was
deemed appropriate considering the immature state of the
technology. At this point, the greatest perceived service the
Standards Committee can provide is to make the KBS life cycle
explicit. In short, a by-the-numbers approach to KBS
development, V&V, fielding, and maintenance.
The first official task the Standards Committee will be to
determine its objective(s). The discussion at the first meeting
suggests an initial focus on the KBS lifecycle. In the next few
weeks the Standards Vice-Chair will call all the initial meeting
attendees and solicit their inputs abaout objectives. Anyone not
attending the meeting who would like to help with this can call
the Standards Vice-Chair, Capt. Dave Howell, during working
hours, at (513) 257-2571.
----- EXPERT SYSTEMS IN GOVERNMENT -----
AISIG 1990
by: Dan Yurman
DYU@NCCIBM1
202-475-6754
I am pleased to announce our participation in the 5th annual "AI
Systems in Government Conference" to be held in Washington, DC,
in May 1990. The Task Force on Expert Systems has agreed to
chair the panels for the 1990 Conference. If you are interested
in convening a panel or being a speaker on one, please contact
one of the panel committee members listed in this article. The
panel committee met on May 10th. Following is the committee
report.
1989 Experience
In 1989 there were seven panels. The bulk of the conference was
taken up by technical sessions spread over two tracks and the
plenary sessions. The conference, including tutorials and
exhibits, lasted five days. Approximately 200 people came to one
or more sessions. I believe there will be better attendance in
1990. This will be due to the fact that there is a determined
effort not to conflict with the AI in Logistics Conference.
We identified more than 7 potential panels. Additionally, Jerry
Feinstein, who chaired panels at the 1989 conference, has joined
us as a working member for 1990 and has agreed to find conveners
for two panels. Following is a list of proposed panels. I see
no reason why we should feel constrained to focus only on these
panel topics if better ones emerge from our respective
discussions with potential conveners. For this reason, I regard
the potential topics listed below as "straw men" until our
meeting in July. As is the case with any conference involving
volunteer speakers, some conveners and panelists may drop out by
May 1990.
Panel Assignments
Dan Yurman, Chair
US EPA (OS-110)
401 M St., SW
Washington, DC 20460
(202) 475-6754
- AI Shells
- AI at EPA
- Software Engineering for Expert Systems
- Medical Applications
Harry Siegel
JAYCOR
1608 Spring Hill Rd.
Vienna, VA 22182
(703) 847-4120
- NASA Expert Systems
- AI for Acquisition/Logistics Management
- SDI Expert Systems
- AI Hardware Platform Developments
Ralph Wachter
Office of Naval Research
Computer Science Dept.; Code 1133
800 N. Quincy St.
Arlington, VA 22217
(703) 696-4304
- Executive Panel on AI
- AI Laboratory Chiefs
- Innovative Technology Programs
Jerry Feinstein
ICF/Phase Linear
9300 Lee Highway
Fairfax, VA 22031
(703) 934-3280
- How the Press Views AI
- International Applications
(via video tape)
Unassigned
- State Government Applications
- Local Government Applications
1990 Plans
Panels in the 1989 conference lasted 1 hour 30 minutes each
except one on SDI which lasted 2 hours. Assuming there are four
or five speakers for a panel, and assuming time is desired for a
Q & A session with the audience, the convener must plan speaking
time for each panelist accordingly. This include a few minutes
to introduce everyone, time between speakers to get organized
with A/V equipment, etc. The point is that time slips by fast
and speakers should be advised to be crisp and to the point in
their talks.
Next Meeting
The next meeting of the panels committee is July 19, 1989, at the
JAYCOR office at 10 AM. At that meeting, the committee will
share the following information about our panels.
- Name of panel convener & biography
- Objectives for the panel
- "Who should attend" information
- Brief bibliography, if possible
Some of this information will come later from the panel conveners
themselves. The full conference committee has requested that
panels be "locked up" with sufficient time to list the names of
panel conveners and speakers in the final program. I will get the
deadlines for this information at a meeting with the committee on
May 30th.
Open Issues
We did not achieve closure on several points. This was due in
most cases to a lack of information. I would like to provide a
summary of our discussions at the May 10th meeting. I hope I
adequately convey your thoughts here, and if not, I regret any
oversight.
1. Focus on Applications
Panels should emphasize concrete applications whenever possible.
People attending the 1989 conference expressed the desire to hear
about applications which added value or which improved
productivity in the organization which used the system.
Alternatively, speakers have the option of providing evaluations,
either quantitative or qualitative, of systems which were built
as prototypes but not shipped to users. This satisfies the
requirement for a focus on "lessons learned."
2. Speakers from Industry
We discussed whether it would be useful to try to draw in
speakers from vendors of computer hardware and software. It was
pointed out that while many computer companies, such as IBM, DEC,
and TI, make heavy use of expert systems inhouse, fewer offerings
have found their way to market. Some firms prefer not to
advertise their use of expert systems since it alerts the
competition to the technology which is used to gain and keep
customers.
In the post-conference evaluation of the 1989 meeting, it was
remarked that exhibitors complained that those attending the
meeting did not have enough time to go to the exhibit floor.
Perhaps a source of speakers would be firms supplying expert
system hardware or software who have government clients with
nonclassified applications. Certainly, the successful use of a
product in a public government setting could be "packaged" to fit
a panel presentation at the conference. Please give this your
consideration.
3. High Quality Handouts
Although the conference publishes a set of proceedings, panel
presentations are not included in it. For this reason, handouts
from speakers are very important.
It was suggested that speakers include a brief bibliography for
further reading on their topic. For instance, the bibliography
could include a standard reference as well as well known journal
articles by experts on the topic. Citations should be accessible
in English in the U.S. A citation format is available from the
IEEE.
It was suggested that speakers provide brief biographies, which
is certainly a benefit. Harry Siegal agreed to prepare a draft
format for speaker bios. He noted that these should be brief
with the objective of placing all speakers on an equal footing.
4. Recognition for Panels
It was suggested that each panelist receive a certificate
suitable for framing or a letter of appreciation for appearing at
the conference. A suggestion was made to award a prize for the
best panel, but there was uncertainty about how to make the award
"on the spot" by the end of the conference. A suggestion was
made to provide "tokens" of appreciation such as a coffee cup,
desk ornament, or other novelty item.