Call for Participation in
Agent Based Social Simulation
SIG Meeting in London
21st and 22nd April, 1999
Topics of interest concern the SIG's further activities in European
R&D of agent-based social simulation, presentation and discussion of
collaborative efforts and advances in R&D in the area. A preliminary
schedule will be available soon.
The meeting will be divided into four sections which are intended to
cover the range of interests of participants. If you feel strongly
that some alternative or additional section should be included, let the
organizers (Rosaria
and Scott) know as soon as possible.
In order that the SIG meeting successfully reflects the interests and needs
of the participants, it is essential that you propose topics within
these sections. Discussion about topics, both to define what is wanted
and to inform the discussions in London will be appropriately conducted
on the ABSS email list.
Please note that this meeting is not an academic seminar but rather
a meeting to explore common interests and potential collaborative projects.
In all sections, a special joint effort is expected and encouraged in (a)
edge cutting projects (application of simulation to new fields or sub-fields),
(b) integrating approaches (e.g., coupling formal/theoretical instruments
with simulation models and results, etc.;) and (c) cross-methodological
research (e.g., comparing data from real phenomena with simulation findings)
A particular subject of discussion will be the establishment of,
or SIG support for, participants' applications for Fifth Framework Funding
and, in particular, collaborations with groups funded outside of the EU
such as NSF-funded American groups.
Attendance at the meeting is
by invitation only. All of the available places have been allocated.
There will, however, be several meetings a year. So even if you missed
the application deadline for this meeting, there will be an early opportunity
to attend the next one.
The programme, updated in light
of participants' comments and scheduling requirements is available here.
Four topic areas are covered:
ABSS for socio-economic, cultural and environmental
development
Agent-based social simulation has a great potential for (a) investigating
several aspects of social complexity: socio-economic (e.g., the trade-off
between growth and welfare), cultural (e.g., diversity, multi-culturality,
evolution and spread of cultural traits, etc.) and environmental (e.g.,
sustainable development) and for (b) contributing to complexity management
(institutional design and policy-making). Under this heading, participants
are invited to focus on how to bridge the gap between different levels
and domains of complexity, namely - the interplay between, and the mutual
impact of cognitive, social and environmental complexity, - connections
between formal/legal and informal/social institutions (norms, conventions,
laws, roles, etc.) - connections between cultural and social dynamics and
evolution.
ABSS for computer-mediated agent societies
The various domains of applications of agent systems highlight the necessity
for a better understanding of competitive and cooperative agent societies
(groupware, elctronic commerce, virtual organizations, horizontal organizations,
etc.). Countless problems are encountered (delegation, privacy, authority,
costs of negotiation, monitoring, control management etc.) and many solutions
are proposed (trust relationshpis, conventions, laws, reputation, coalition
formation, distributed monitoring, delegated control, etc.). Most solutions
are based upon deontic notions and refer to phenomena traditionally explored
in other fields of science (social sciences, logical philosophy, deontic
logics, philosophy of law, etc.). Participants are invited to present projects
on the use of agent-based social simulation for understanding and managing
computer-mediated agent societies. Again, a special consideration will
be reserved to projects combining formal/theoretical work and simulation
methods and techniques.
ABSS and agent-systems
Whilst simulation techniques in the software agents field are aimed at
checking systems efficiency, the use of agent-based simulation in the social
sciences is aimed at building and exploring a virtual world, thereby implementing
and integrating models of its different components (of the world itself,
of the agents and of other objects within it). Therefore, (a) agent-based
social simulation has a great potential for several software agents sub-fields
(MAS, believable agents, synthetic actors and personalities, etc.) and
the related domains of application (entertainment, animation, tutoring,
assistance, etc) which need a better appreciation of social systemic effects,
of social complexity, and of social situated-ness. On the other hand, agent
systems have a great potential for updating the models and architecture
of the agent used in social simulation. Participants are invited to present
projects where these approaches are combined, evaluated and compared.
Software for ABSS
There are two widely distributed software modelling environments for social
simulation: Swarm
and SDML. The merits
of each have recently been discussed on the simsoc list. In addition,
several social simulation teams have been developing bespoke enviroments
that are less widely distributed. Many in the social simulation community
find Swarm and SDML difficult to learn, It seems that the most able
in this area are Ph.D. students. A discussion about the needs of
the social simulation community and means of meeting those needs will help
individual developers to provide better software and, perhaps, facilitate
replication studies and the use of one another's models.
Rosaria Conte
Division of AI, Cognitive and Interaction Modelling
PSS (Project on Social Simulation)
IP/Cnr
Viale Marx 15
00137 Roma
ITALY
voice: +39+06+86090210
fax: +39+06+86090214
email: rosaria@pscs2.irmkant.rm.cnr.it
http://pscs2.irmkant.rm.cnr.it/users/rosaria/home.html |
Scott Moss
Centre for Policy Modelling
Manchester Metropolitan University
Aytoun Building
Manchester M1 3GH
UK
voice: +44+(0)161+2473886
fax: +44+(0)161+2476802
email: s.moss@mmu.ac.uk
http://cfpm.org/~scott |
Last modification: 29 March 1999, 1530 UTC by Scott
Moss
ABSS SIG Homepage