Automated Dynamic Audit Programme Tailoring: An Expert System Approach.

Abstract This paper describes the development of ADAPT, an expert system audit automation tool designed to assist the auditor in tailoring audit programmes. ADAPT deals with the aggregation of audit assurance within an explicit model of the structure of audit evidence based on financial statement assertions. It is not clear that such a project, with an operational expert system as its main objective, can ever be said to be completely finished. However, after three years of development, ADAPT has recently been released for field use. The feedback from limited field use this season will be used to develop the software facilities further. Meanwhile, as described in the paper, the software incorporates numerous facilities for the exploration of modified models of the aggregation process. This draft paper is based on an estimated 5000 pages of various types of documentation generated during the project, some of which are still being worked on. The objective of producing a summary at this time is to take the opportunity to comment on how research exposed at the USC Audit Judgment Symposium in recent years has influenced the development of a practical audit support tool, while the project itself has generated additional research issues. At this early stage, time has not yet permitted this draft of the paper to describe all the related work covered in the literature. However, references are given at the end of the paper to the main articles and publications that the developers have found helpful. Most of these references are not cited specifically in the body of the paper, but interested readers may contact the author directly.1 This is a shortened version of the paper presented at the USC Audit Judgment Symposium; the longer version includes additional material describing the user interface, system development, assertions, the calculation of reliability, a variety of different aggregation rules, the Control Strategy, heuristics and knowledge engineering issues in greater detail.