Knowledge Representation Theory and the Design of Auditable Office Information Systems.

Intelligent Office Information Systems (OIS) integrate knowledge representation, natural language processing, information management, problem solving and other capabilities intended to help attain organization goals. Organizational control of resources is sometimes given less attention by OIS developers than thought warranted by internal and external auditors. This paper develops a knowledge representation scheme based on natural language processing concepts in support of dynamic, cooperative organization problem solving which, at the same time, assures the design of a controlled and auditabte OIS. The representation scheme and theory are based on Discourse Representation Theory, Conceptual Graph Theory and Scripts. Queries involving direct data retrieval, post- and precondition analysis and deduction are processed by the system. An example based on an accounting internal control system is included in the paper.