A. Salah (USA)
Knowledge Engineering, Formal Model, Agent-basedSoftware, Decision theory
Intelligent agents are encapsulated software systems that are situated in some environment and capable of flexible, autonomous action in that environment in order to meet its design objective. The approach of building complex software systems based on agent technology is emerging lately, that is agent-based software. Agent-based software engineering is a scientific approach supports the handling of complex software by assuming agents as components in complex software. One of the major problems in the agent-based approach is how to specify, verify and test agents in a rigorous way compared to traditional software. Agent specification and testing has to be built on an abstract level and verified, rather than testing particular cases. This paper introduces a modeling technique and a formal representation for knowledge and decisions based on integration of decision tables theory and techniques as a model for decision-making in agent-oriented software. This modeling technique adds an abstract level through which decomposition can be achieved systematically. The formal theory of decision tables provides a scientific basis for the specification and verification for agent design and maintenance. It is shown through the paper that this modeling technique satisfies modularity, semantic characteristics, comprehensibility, hierarchical organization, and verifiability. The modeling technique is on an abstract level independent from the implementation technique.
Important Links:
Go Back
