KNOWLEDGE REPRESENTATION FOR INTELLIGENT ACCOUNTING MACHINES
Abstract
This research proposes a knowledge representation for intelligent accounting machines. An intelligent accounting machine is operationally defined as a machine capable of accepting and processing accounting information and responding to queries in a manner consistent with generally accepted accounting principles. The machine consists of an inference mechanism which operates on a knowledge base containing both accounting data and operations defined on the data. The knowledge representation may be logically divided into three segments. Data about economic events and objects is represented in a relational format. Accounting principles and procedures are represented via axioms and procedural attachments. Natural language recognition knowledge is represented axiomatically via definite clause grammars. This architecture allows the inference mechanism to integrate various types of knowledge for processing while maintaining independence between them as much as possible. Advantages of this approach to accounting system design include improvements in auditability, portability, extensibility, and user interface, as compared to traditionally designed systems.
Degree
Ph.D.
Subject Area
Management
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.