Multi-Agent Architecture for Knowledge-Driven Decision Support - Distributed Artificial Intelligence, Agent Technology, and Collaborative Applications - page 30

The implementation incorporates XML related technologies for knowledge representation, storage and
knowledge exchange among participating intelligent agents to deliver decision support to the user. In
IKMDSA agents provide distributed intelligent decision support by exchanging their knowledge using
XML and its related set of standards. Implementation details of the architecture and implications for
further research in this area by academics and practitioners are provided.
In section 2, we review relevant literature in intelligent agents and the role of decision trees in induc-
tive learning and knowledge representation in terms of decision rules. In section 3, we discuss the role
of XML in representing and facilitating knowledge exchange for intelligent agents. Section 4 provides a
detailed description of the various components of the IKMDSA architecture and their inter-relationships
in facilitating the creation, representation, exchange and use of domain specific knowledge for decision
support tasks. In section 5, we provide a detailed description of the implementation of the architecture
through the use of an illustrative example. Section 6 includes a discussion of the implications of integrat-
ing KMS and DSS support in business, and the role of the proposed IKMDSA architecture. Section 7
concludes with limitations and future research directions.

LITer ATure r eVIeW

Knowledge is an important organizational asset for sustainable competitive advantage. Organizations
are increasingly interested in knowledge-driven decision analytics to improve decision quality and the
decision support environment. This requires use of corporate data to develop higher-level knowledge in
conjunction with analytical tools to support knowledge-driven analysis of business problems (Ba et al.,
1997). Advances in systems support for problem solving and decision-making increasingly use artificial
intelligence (AI) based techniques for knowledge representation (KR) (Whinston, 1997; Goul et al.,
1992; Goul and Corral, 2005). KR takes multiple forms including business rules, decision analytics and
business intelligence generated from various machine learning algorithms and data mining techniques.
Intelligence is the ability to act appropriately in an uncertain environment to increase the probability
of success and achieve goals (Albus, 1991). Intelligent decision support systems (IDSS) incorporate
intelligence as problem domain knowledge with knowledge representation that informs and supports
the decision process to facilitate problem solving and reduce the cognitive load of the decision maker.

Software Agents and Intelligent decision Support Systems (IdSS)

An intelligent agent is “a computer system situated in some environment and that is capable of flexible
autonomous action in this environment in order to meet its design objectives” (Jennings and Wooldridge,
1998). The terms agents, software agents and intelligent agents are often used interchangeably in the
literature. However, all agents do not necessarily have to be intelligent. Jennings and Wooldridge (Jen-
nings and Wooldridge, 1998) observe that agent-based systems are not necessarily intelligent, and require
that an agent be flexible to be considered intelligent. Such flexibility in intelligent agent based systems
requires that the agents should be: (Bradshaw, 1997; Jennings and Wooldridge, 1998)

• Cognizant of their environment and be responsive to changes therein

• Reactive and proactive to opportunities in their environment

• Autonomous in goal-directed behavior