ABSTRACT: Efficiency considerations aside, an information system may be regarded as a set of logical axioms about the organization and its environment. To what extent are these axioms shared between applications? Between organizations? Within an industry? A knowledge engineering approach to systems analysis is suggested whereby (prototype) systems are built by adding refinements to a set of background axioms representing knowledge about business, administrative functions, etc. New applications may thus build on the knowledge from previous applications, rather than starting from scratch. The emerging technology of logic programming provides a suitable implementation language for this approach. However, logic programming by itself is not enough; extendibility and probability of axioms require a consistent semantic basis across applications. Semantic issues are discussed, and a specific ontology is proposed.
Key words and phrases: knowledge-based systems, MIS semantics, logic programming, logic databases