The most important lesson of this chapter - and perhaps one of the most important lessons of multiagent systems generally - is that when faced with what appears to be a multiagent domain, it is critically important to understand the type of interaction that takes place between the agents. To see what I mean by this, let us start with some notation.
Chapter 10 give an overview of work that has been carried out on the development of methodologies for multiagent systems. This work is, at the time of writing, rather tentative - not much experience has yet been gained with them. This chapter begin by considering some of the domain attributes that indicate the appropriateness of an agent-based solution. I then go on to describe various prototypical methodologies, and discuss some of the pitfalls associated with agent-oriented development.
Agents have found application in many domains: in this chapter, I will describe some of the most notable. Broadly speaking, applications of agents can be divided into two main groups. Distributed systems: In which agents become processing nodes in a distributed system, the emphasis in such systems is on the 'multi' aspect of multiagent systems. Personal software assistants: In which agents play the role of proactive assistants to users working with some application. The emphasis here is on Individual' agents.
The 'traditional' approach to building artificially intelligent systems, known as symbolic AI, suggests that intelligent behaviour can be generated in a system by giving that system a symbolic representation of its environment and its desired behaviour, and syntactically manipulating this representation. This chapter focus on the apotheosis of this tradition, in which these symbolic representations are logical formulae, and the syntactic manipulation corresponds to logical deduction, or theorem-proving.
In the three preceding chapters, we have looked at the basic theoretical principles of multiagent encounters and the properties of such encounters. We have also seen how agents might reach agreements in encounters with other agents, and looked at languages that agents might use to communicate with one another. So far, however, we have seen nothing of how agents can work together. In this chapter, we rectify this.