An investigation into applying UML to the Zachman framework

The Zachman framework is considered to be the most referenced framework for the purpose of enterprise architecture. It is commonplace to compare other frameworks with this basic one in order to show correctness and usability of those frameworks. However, this is more than a fashion, the Zachman framework is actually the best one. Despite of its popularity, the Zachman framework could be a challengeable one in practical situations because there are not enough well-known methods and tools covering all of its aspects. Three major challenges in using this framework, are discussed in this article. These challenges are lack of a methodology, a well-defined repository and a popular modeling notation. Focus of this article is on solving the last problem with the help of notations in UML (Unified Modeling Language) and UML Business Profile. At the first glance the topic seems to be already researched by others, but there are some major distinctions between this work and the others’, which make it a unique one. Most of the other work tried to cover the framework using multiple class diagrams stereotyped in different ways. This work tries to cover the Zachman framework using all of the UML features, especially those, which are convenient in common modeling tools as well as ignoring unfamiliar symobls as it is used by some authors. A case study is used upon which we show how to apply the selected notation on a sample enterprise to develop cells in second and third rows of the framework. Models are tested to consider if they are supporting Zachman rules governing the framework. Furthermore, in order to see if they could be convincing enough, a statistical study is employed. Although A. Fatolahi ( ) · F. Shams Electrical and Computer Engineering Faculty, Shahid Beheshti University, Post Box 15875/1761, Velenjak, Tehran, Iran e-mail: Ali [email protected] results of these tests are relatively acceptable, the problem of inventing new modeling notations is mentioned as an open problem.