Abstract:
This dissertation describes the development and structure of a rule-based knowledge-based system that assists users with the development of models used to support the decision making process. The characteristics of decision making, decision support systems, modelling methodologies, and those problem characteristics suitable for processing by knowledge-based systems are compared. The outcome of the comparison is discussed leading to the conclusion that a rule-based knowledge-based system is suitable for assisting decision, makers with the construction of models used to support them with decision making. Processing concepts used by MODRULE are described and include an explanation of: model statements, model rules, entity matching, the inference process, the control mechanism, and the construction of reduced rule bases. Implementation details and the logical structure of MODRULE is outlined, followed by typical examples of model extension, generation, and execution. Results of model extension, generation and execution are discussed, followed by conclusions.