Integration of an Accounting Domain Ontology (REA) with an Upper Ontology (SUMO) invited talk by Professor Bill McCarthy (Michigan State University) on 03/17/2005

* Subject/Agenda Professor William E. McCarthy will be giving a talk entitled: andquot;The Integration of an Accounting Domain Ontology (REA) with an Upper Ontology (SUMO)andquot; * Date Thursday, Mar. 17, 2005 * ONTOLOG forum Wiki page details http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2005_03_17 * Abstract The REA (Resource-Event-Agent) enterprise ontology is a well-accepted model of the concepts and relationships that exist in a normal economic exchange between companies or conversions within companies. REA is based upon a foundation grounded in both accounting and economic theories of the firm, and it is used in the accounting domain extensively for both research and teaching in accounting systems. The seminal REA paper was published in The Accounting Review in 1982, and its basic framework has been extended multiple times in recent years in work by Geerts and McCarthy. The REA ontology has been used in standards work within the UN/CEFACT, ISO, and other standards bodies. All domain ontologies need to reconciled eventually to an upper ontology if they expect to be used extensively across domains, and the interrelated domain concepts and associations need to be mapped to those of the upper ontology where those correspondences exist. The upper ontology to be used in this mapping exercise with REA is SUMO, the Suggested Upper Merged Ontology, one of the IEEE Standard Upper Ontology working group starter documents. SUMO is extensively used and researched in the computer science community, and its conceptual structures are better defined and integrated than those of more informal ontological frameworks like Bunge and Sowa. Sowa's conceptual framework fits REA well, but Bunge's is a clear conceptual mismatch. Mapping to more general ideas and axioms can be both a quality control and a concept expansion exercise. There are bound to be many gaps and overlaps, just as one would find in mapping individual external schemas to a more general conceptual schema in the database design process of view modeling and view integration. The integration exercise to be described in this presentation is most certainly preliminary, and it is being accomplished by an REA expert who is also an acknowledged SUMO novice. The domain concepts of REA work extremely well in the process modeling of inter-firm business collaborations and in the re-orientation of basic accounting ideas away from the artifactual grip of general ledgers and traditional accounting reporting. Their integration with SUMO will be another step on the path to assessing their suitability for use as a more general enterprise ontology.