The Analysis and Design of Embedded Knowledge-Based Systems Using Box Structure Methods

Abstract

Abstract:Knowledge-based systems are widely used for decision support system and expert system applications. However, they are not a panacea for all business problems. Thorough requirements analysis methods must determine the need for knowledge-based systems vis-a-vis alternative information system designs. Furthermore, knowledge-based systems are often used as embedded subsystems within larger organizational information systems. Rigorous design methods must support the development of complex information systems with integrated, heterogeneous subsystems, including knowledge-based subsystems. In this paper, we demonstrate that the box structure methods of information systems development support both the requirements analysis of knowledge-based systems and the design of embedded knowledgebased systems. An extended example is presented to show the use of the box structure methods in the development of a financial planning system that requires embedded knowledge-based system components.