Newsgroups: comp.ai
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!Germany.EU.net!EU.net!news.sprintlink.net!cs.utexas.edu!math.ohio-state.edu!jussieu.fr!univ-lyon1.fr!swidir.switch.ch!scsing.switch.ch!gdvdns1!gdv015!liver
From: liver@vptt.ch (Liver Beat fe52)
Subject: IJCAI-95 Workshop on AI in Distributed Information Networks
Message-ID: <D4v5Dz.314@vptt.ch>
Sender: root@vptt.ch (Operator)
Nntp-Posting-Host: gdv015.vptt.ch
Reply-To: liver@vptt.ch
Organization: Swiss PTT
Date: Fri, 3 Mar 1995 12:01:11 GMT
Lines: 186


-------------------PLEASE CIRCULATE AS WIDELY AS POSSIBLE---------------
		  

	*** LAST CALL FOR SHORT POSITION PAPERS ***


     IJCAI-95, Montreal (Canada), August 19, 1995

    Workshop on AI in Distributed Information Networks,

Where are the successes and where are the challenges for the future?


Objective
---------
Distributed Information Networks (DIN) deliver information over a variety of
networks which include telecommunication, cable, wireless and satellite
networks. This workshop will address the past and the future of AI in
information networks. Two classes of problems have traditionally been
addressed by AI for these networks. The first class of problems has been 
in network management which addresses diagnosis, monitoring, alarm
filtering, routing and traffic analysis. The second class of problems
has been in applying AI to manage and present the information flowing
along networks.  Information network applications are commonly referred
to as services by the network providers. Current services for end users
vary from information filtering and financial services available through 
personal computers with modems to talking yellow pages available through 
telephone switching centers and distributed databases.

Both classes of problems have had successes and failures with the use
of AI. One objective of the workshop will be to discuss and categorize
the results of using AI in both network management and information
network applications. A key question to address is **why has AI succeeded
or failed for various DIN problems**?

High speed heterogeneous networks formed by integrating telephone,
cable, wireless, satellite, etc. networks pose a variety of new
problems in network management. They also present new opportunities in
information network applications such as personalized electronic
newspapers, video-on-demand, personalized home shopping, and nomadic
computing with personalized digital assistants to name a few. The
second objective of the workshop will be to discuss and enumerate the
challenges for AI in managing the new integrated networks and the
opportunities in making use of them. A key question to address is 
**what are the problems and what AI approach (if any) can be envisioned 
as part of the solution**?

Short (2-3 page) position papers submitted **e-mail text or PostScript**
only are sought on: 
1) AI in network management, 
2) AI in information applications, 
3) AI *visions* for network management and network information applications, & 

Specifically, *short position papers* are sought that outline 
successes or failures in applying AI in a novel manner to:
a) network management problems such as, but not limited to
- fault and/or alarm management
- performance monitoring
- configuration
- adaptive routing
- diagnosis
- distributed intelligent agents
- scheduling
- provisioning

b) information network applications such as, but not limited to
- information filtering (including off the world wide web)
- digital libraries
- personal digital assistant applications
- video-on-demand
- telemedicine
- distance learning
- talking yellow pages

SHORT POSITION PAPERS that outline a well-supported vision as to where 
the challenges lie for AI in the new networks are sought.


Proposed Schedule
-----------------

Short Position Papers received:  March 13th,      1995
Author notification:             March 20th,      1995
Final papers distributed:        June 1st,        1995
Begin e-mail discussions:        June 15th,       1995
Workshop:                        August 19th,     1995
Call for book chapters:          September 19th,  1995
Revised papers:                  December 19th,   1995
Submission to a Publisher:       December 30th,   1995

Short Position Paper submissions or long abstracts (limited to 3 pages)
will be used as the basis of invitations for people who want to
participate in the discussion at the workshop without having submitted
full papers. Short Papers must include: author's name(s), affiliation,
complete mailing address, phone number, fax number, and e-mail.
Text-only or Postscript submissions should be sent to the chairperson
by e-mail at suhayya@ai.iit.nrc.ca or by anonymous ftp, contact the
chair for more information re-ftp info.

The goal is to encourage participants to take a position as to where AI
has succeeded or failed in managing networks and information network
applications and where the challenges lie for the future. The
participants are then expected to defend their position at the
workshop. The workshop will be limited to 40 active participants.
Additional information will be available from the chair by request.
Attendees **must** register for the main IJCAI-95 conference.


Day of the Workshop Organization
--------------------------------
The goal of the workshop is to maximize discussion and avoid
having a miniconference. To meet this goal, the workshop will 
be organized as two sets of panels, 1 morning and 1 afternoon:
o successes/failures of AI in information networks
o vision for embedded AI technologies in information networks

The authors of the best papers addressing one of the panel topics
will be chosen for the panel. The panelists will be given 5 minutes to
present a position on the topic and the workshop participants will
discuss and question the panelists on the topic. The panelists and
respective moderator will be given 30 minutes to introduce the topic
and 45 minutes will be scheduled for discussion of each panel topic.
The moderators will use some of the e-mail discussions to guide them in
focussing the panel discussion. The goal of the two morning panels will
be to enumerate the AI techniques or approaches that have succeeded or
failed in the particular class of problem. The goal of the afternoon
panels will be to look ahead and identify the roles that AI can play in
the management and the applications of the new networks. Four
appropriate Committee members will be asked to moderate the panels to
ensure that the discussion is lively and useful. The chair will ensure
that the panelists address their topics, stay within their alloted time
and she will summarize the areas for a group discussion that will
conclude the workshop.

Organization of the Work
-------------------------
As described in the proposed schedule, the workshop will begin through 
discussions by electronic mail 2 months before the one-day workshop.  
This strategy has been used successfully used for workshops at IJCAI-93 
and AAAI-94. From the accepted papers, two discussion themes will be pursued: 
the success and failure of AI in the past, and the challenge for AI in the
new networks.  After the one-day workshop, longer versions of the best
papers will be solicited and submitted for publication in a book on AI
and Distributed Information Networks: past and future. In addition, a
paper summarizing the results of the workshop will be submitted to a
widely circulated publication such as IEEE Expert or AI Magazine.


Organization Committee
----------------------

Chair
-----
Sue Abu-Hakima, Knowledge Systems Lab, National Research Council, Ottawa, Canada

Committee
---------
Stephen Corley, Intelligent Systems Unit, BT Labs, Suffolk, UK
Michel Feret, SuperNode Tools, Bell-Northern Research, Canada
Andreas Girgensohn, NYNEX Science & Technology, White Plains, NY, USA
Nancy Griffeth, Bellcore, Morristown, NJ, USA
Jacob Groote, PTT Research, Groningen, Netherlands
Yezdi Lashkari, MIT Media Laboratory, MIT, Cambridge, MA, USA 
Beat Liver, R&D, Swiss Telecom PTT, Berne, Switzerland
Peter Selfridge, R&D, AT&T Bell Laboratories, Murray Hill, NJ, USA
Dominique Sidou, Institut Eurecom, Sophia Antipolis, France
Jon Wright, AT&T Bell Laboratories, Murray Hill, NJ, USA

Main contact address of the Organizing Committee:
-------------------------------------------------
Sue Abu-Hakima					
Knowledge Systems Lab
Institute for Information Technology
National Research Council of Canada
Building M-50, Montreal Road
Ottawa, Ontario K1A 0R6
tel: (613) 991-1231
fax: (613) 952-7151
email: suhayya@ ai.iit.nrc.ca

------------------------------------//--------------------------




