A Problem Representation Approach for Decision Support Systems

The choice to include the human in the decision process affects four key areas of system design: problem representation, system analysis and design, solution technique selection, and interface requirements specification. We introduce a design methodology that captures the necessary choices associated with each of these areas. In particular we show how this methodology is applied to the problem representation process for an actual decision support system for satellite operations scheduling. We highlight the main issues related to problem representation in the large and how these issues seek to capitalize on the strengths of the computer system and the human user. We discuss our implementation of a scheduling problem representation scheme that embodies these concerns and we show the decision support advantages gained as a result of this representation.