From: Daniel Kudenko (kudenko@cs.york.ac.uk)
Date: Thu 10 Oct 2002 - 15:59:03 BST
Dear all,
This email introduces the ALAD SIG and outlines current and future
activities. I apologize for the length of this email, but I hope that it
will give you a good overview and encourage you to participate in ALAD.
As most of you already know, ALAD is an Inter-network Special Interest
Group on Agents that Learn, Adapt and Discover. ALAD is part of
AgentLink II and co-sponsered by EUNITE and ILPNet II. Also, as you may
be aware, there has recently been a change in coordination of this SIG:
it is now coordinated by me (Daniel Kudenko), Eduardo Alonso, and
Dimitar Kazakov.
To date, much learning agent research is being done in a wide range of
application domains and contexts, and we believe that the field could
benefit from closer contacts between the research groups and the
resulting exchange, discussion, and generation of new ideas. We hope
that ALAD SIG will become a useful platform to achieve these goals.
In this context, we plan the following activities:
1. Scientific Meetings
The next AgentLink SIG meetings will take place on Feb. 3rd and 4th in
Barcelona. We are currently planning this event, and suggestions are
more than welcome. Please watch out for a call for participation soon.
One of the highlights will be the I2A/ALAD joint session on learning
information agents.
Also, we plan to organize a third Symposium on Adaptive and Learning
Agents, following up on the previous successful Symposium Series (see
http://www.aamas.net). The Symposium will probably take place as part of
the AISB'03 convention in Aberystwyth in April 2003. A call for papers
will be sent out soon.
2. ALAD Web Page
The ALAD web page is located at http://www.cs.york.ac.uk/~kudenko. We
envisage this page to become a major research resource for the area of
learning agents. The page is currently under construction, and please
have a look at it, and let me know with any comments, suggestions, or
additions.
3. Matchmaking
Given the broad and multi-disciplinary character of the area of learning
agents, many research efforts can benefit from wider collaboration. We
will try to provide a resource page (as part of the ALAD web pages) that
will help finding research partners for proposals and ongoing projects.
4. Publications
We plan to produce proceedings as well as books on the topic of learning
agents. A Springer volume on this topic is currently being edited by us,
and we hope that further volumes will follow.
5. Roadmap Document
AgentLink is currently producing a roadmap document for teh EU, and a
draft will be available soon for comments. We have already been involved
in the ALAD-related content, and we hope to get more feedback from the
SIG members once the first draft is produced.
In addition, we believe that it would be of great benefit to the
learning agent research community, if ALAD would produce a separate
roadmap document for the EU and other funding bodies, that outlines
important areas of future research. More on this activity later ...
6. Definition of benchmark tasks
To date, there are many learning agent techniques available, but often
only little is known about their scalability to real-world tasks, and
comparisons to other methods are often difficult. To facilitate a more
unified study and encourage researchers to work on real-world domains,
ALAD could produce benchmark tasks and domains for learning agent
technologies. We specifically encourage industrial researchers and
practitioners to be involved in this activity, to ensure realism and
practical applicability of these benchmark tasks.
Phew, that's been quite a lot for the first ALAD email, and I hope that
it will spark off a discussion. Please do comment on any of the above,
and send suggestions for further ALAD activities. We need your input to
make all this work.
Best wishes,
Daniel
--
Dr. Daniel Kudenko Office: CS202B
Department of Computer Science Email: kudenko@cs.york.ac.uk
University of York http://www.cs.york.ac.uk/~kudenko
Heslington, York YO10 5DD Tel: +44-1904-434776
United Kingdom Fax: +44-1904-432767