With the advent of e-systems; business and consumers have access to a plethora of information which makes the decision making process more complex. The overwhelming information flow makes it extremely difficult for decision makers to analyze the available data and make precise decisions. Under such an information intensive online environment, businesses need to make real time intelligent decisions in order to stay economically and commercially viable. Multi agent systems have the inherent ability to facilitate provision of an adequate decision support mechanism in an e-business setting. These multi –agent decision support systems are of particular assistance in processing large amounts of data, filtering out irrelevant information and eliminating cognitive biases. In this study, we attempt to explore the existing decision support mechanisms facilitating the e-business environment with respect to B2B and B2C segments. Several models have been proposed and implemented to assist decision support in e-systems. We attempt to identify possible weaknesses in these models and pin point future research areas that would provide an opportunity for improved decision support mechanisms in e-systems.
1.1 Agents, Distributed Artificial Intelligence and Multi-Agents
With the advent of computer systems, there has been a significant progress in terms of framing online transactions processing systems. These systems are programmed to carry out specific activities which are restricted to facilitation of online transaction processing only and do not assist in the decision making process as they lack the “intelligence” which is the basic essence of the modern decision support systems. OLTP systems only help automate the transactions so as to enable easy record keeping and limited human involvement. Typical example of an OLTP system is the inventory management system, accounting system or similar database systems that might be in use in different departments of an organization.
Computer systems that have the inherent capacity to take decisions when required are referred to as agents. These systems are designed to make intelligent choices according to the situation. These agents when operate in a rapidly changing environment where real time decision making is required are known as intelligent agents. Typical examples of a multi-agent system include a computer program that would perform an internet search for a typical query. The agent would synthesize the relevant information from various web sources and return a document that would carry relevant information based on our query (Weiss, 1999).
Based on the prelude, we may define an agent as “An autonomous, reactive, pro-active computer system, typically with a central locus of control that is at least able to communicate with other agents via some kind of communication language” (Weiss, 1999)
1.2 Decision Support Systems and Web Based DSS
Decision support systems on the...