Community Computing - Advanced Design Approaches to Emerging Software Systems - page 213

agent systems by modeling the interac-
tions between agents based on the concept
of roles and describing such roles using
an XML-based notation, XRole. Authors
implemented Rolesystem as an interaction
infrastructure of BRAIN, but they did not
concern themselves with cooperation.

tions, multi-agent systems are frequently
developed (Wooldridge, 2000). To coop-
erate with other agents, an organization is
constructed, and each role for performing
a cooperative protocol is dynamically as-
signed to a member agent. The coopera-
tion procedure is able to be predefined or
dynamically determined depending on the
agent's intelligence, but it has not been se-
riously studied.

Cooperative System
Development Approaches

• Community in PICO: PICO is a middle-

ware framework for dynamically creat-
ing mission-oriented communities of per-
vasive objects. To describe structures of
cooperating organizations, it employed
the community concept. It is very similar
to our community computing, but PICO
leaves room for the life cycle of communi-
ty, member specification, the way to assign
software objects dynamically into actual
objects, and the cooperation method.

Cooperation has been a good way to solve a prob-
lem requiring diverse resource and capabilities
and to perform a highly resource-consuming and
time-consuming task (Wooldridge, 1999). The
pervasive service is one of the domains having
such problems, so cooperation has been used in
some research. However, meanings and/or style
of cooperation are slightly different between re-
search studies. In this section, we inspect existing
research studies and discriminate between them
in terms of cooperation. • Community Computing in Digital

Tokyo Project: The authors introduced
community computing in 1998 in order to
support the process of organizing diverse
and amorphous people who are willing to
share knowledge and experiences (Ishida,
1998). The objective of their work is to
make a city-scale supporting system to as-
sist a person's everyday life. In the Digital
Tokyo project (Besselaar, 2002), a commu-
nity is a digitalized representation of real
human communities. All human members
in a community share their preferences and
knowledge and generate a consensus, and
the community computing supports those
processes.

• Team in Computer Supported

Cooperative Work (CSCW): The major
objective of CSCW is to develop group-
ware that effectively performs a common
task using information sharing among all
users (Wilson, 1991; Borghoff, 2000).
Typically, a group in CSCW is a small proj-
ect-oriented team, and a team is defined as
a set of predefined people. Members of a
team are human, and their works are tightly
coupled by sharing information about team
membership as well as information about
the skills and roles of the other members
(Johansen, 1998). In the group protocol
component, the ways in which team mem-
bers cooperate and communicate with each
other are described. Typical groupware of
CSCW are video conferencing systems
and joint document editing systems.

• Community Computing in Microsoft:

In 2005, Microsoft introduced its vision of
community computing (Blau, 2005). It de-
fines community computing as an emerging
technological environment where devices
sharing computing capacities of others and
users' identities are widespread. It insisted

• Group in Multi-agent System: To pro-

vide services requiring complex interac-