Ontological Rules for UML-Based Conceptual Modeling: Design Considerations and a Prototype Implementation

  • Authors:
  • Jeffrey Parsons;Shan Lu

  • Affiliations:
  • Memorial University of Newfoundland, Canada;Around America Aviation Group, Canada

  • Venue:
  • International Journal of Information System Modeling and Design
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

UML is used as a language for object-oriented software design, and as a language for conceptual modeling of applications domains. Given the differences between these purposes, UML's origins in software engineering might limit its appropriateness for conceptual modeling. In this context, Evermann and Wand have proposed a set of well-defined ontological rules to constrain the construction of UML diagrams to reflect underlying ontological assumptions about the real world. The authors extend their work using a design research approach that examines these rules by studying the consequences of integrating them into a UML CASE tool. The paper demonstrates how design insights from incorporating theory-based modeling rules in a software artifact can be used to shed light on the rules themselves. In particular, the authors distinguish four categories of rules for implementation purposes, reflecting the relative importance of different rules and the degree of flexibility available in enforcing them. They propose distinct implementation strategies that correspond to these four rule categories and identify some redundant rules as well as some rules that cannot be implemented without changing the UML specification. The rules are implemented in an open-source UML CASE tool.