REA Economic Agents as Homines Economici in an Agent-Oriented Information Systems Environment

William E. McCarthy
Email: mccarth4@msu.edu
http://www.williamemccarthy.com
Michigan State University
East Lansing, MI., USA

The REA model is a pattern for an internal business process or an external market exchange. It consists of two mirror image constellations of the economic objects involved in a transaction -- economic Resources, economic Events, and economic Agents – connected by a duality relationship.In microeconomic terms, the full REA pattern is a representation of a production function that transforms one or more types of input into an output of greater value. More recent REA work has extended this basic pattern in two directions:

  1. horizontally into types and commitments; and 
  2. vertically into enterprise value chains and workflow representation.
In REA, economic agents are assumed to be homines economici, that is rational parties who engage in arms-length commerce with the specific objective of maximizing utility. For a traditional information system, these REA economic agents are representation artifacts (like tuples in a relational database) that model retrospectively the actual behavior of real world parties. However, in certain types of agent-oriented information systems, the software agents become the economic actors themselves, not mere retrospective representation artifacts. This speech will overview current REA research that seeks to build information system prototypes with intelligent software agents that act as homines economici, both in the way their assumptions, goals, and plans are modeled and in the way in which they act on those modeled components as parties in e-commerce.
 
William E. McCarthy is presently appointed as the Arthur Andersen Alumni Professor of Accounting and Information Systems at Michigan State University.  Professor McCarthy's research and practice interests center around the application of knowledge-based systems, object-oriented analysis patterns, and database theories to the problems of building enterprise information architectures that work within and between firms. His 1982 paper on REA accounting systems was given the first Seminal Contribution to Accounting Information Systems Literature Award in 1996, and he has continued that progress with REA business process patterns in a large number of papers in both business and computer science journals. His most recent work involves adapting REA ontologies to the needs of XML and to the problems of constructing agent-oriented business systems.  REA research working papers are available online and via email request.