Date of Award
Doctor of Philosophy (PhD)
Information Systems and Decision Sciences (Business Administration)
A model management system (MMS) is a computer-based system which facilitates the management of a wide range of decision support models (DSMs). It is a vital component of a decision support system which helps decision makers using data and DSMs in an integrated fashion. It is desirable that an MMS is knowledge-based, flexible, independent and reflecting users' view of a DSM. In addition, an MMS must manage DSMs at both the macro and micro levels. At the macro level, DSMs are described using descriptive attributes to supply users with general problem-solving capabilities. At the micro level, the components and mathematical structure of a DSM are expressed to facilitate DSM formulation. There are three two-level MMS approaches proposed in the literature. All three approaches suggest describing the details of a DSM using the frame system. However, there are problems using the frame system at the micro level. First, the frame system does not provide facilities for handling data of large volume. Second, there is neither design methodology nor evaluation criteria regarding the design of the frame system. Third, pre-defined frame constructs may not be able to encompass all DSMs used in a dynamic environment. Finally, designers of the frame system usually lack of the professional skills to design an appropriate structure. The functional MMS is proposed to overcome the deficiency. It is intended to provide the two-level model management capability with all the desirable features. The macro-level functional MMS is based on first-order logic which is the best-developed knowledge representation methodology so far. At the micro level, the foundation is on relational theory which has proven its usefulness in data management. Additionally, the definitional system used to describe the details of a DSM provides a natural way of developing a DSM in a hierarchical manner.
Ko, Lijen, "A Functional Approach to Model Management." (1990). LSU Historical Dissertations and Theses. 4926.