When developing information systems, data modeling represents a key aspect of the requirements analysis phase. The criticality of the data model stems from the fact that the characteristics of the data govern all other aspects of an information system. As Hoffer, George, and Valacich (2011) state, “data, not processes, are the most complex aspects of many modern information systems and hence require a central role in structuring system requirements” (Hoffer, George, & Valacich, 2011). There are two primary data modeling methods: Entity-Relationship (E-R) diagramming and class-based (usually UML) diagramming. Given the pivotal role that the data model plays in requirements analysis and the success of an information system overall, it is disappointing to find that the two most common data modeling techniques are insufficient to the task. Kroenke and Gray (2006) provide a set of evaluation criteria for data modeling methods and review E-R and UML data models according to the proposed criteria (Kroenke & Gray, 2006). An examination of Kroenke and Gray’s article, “Toward a Next Generation Data Modeling Facility: Neither the Entity-Relationship Model nor UML Meet the Need,” provides an understanding of the state of data modeling and the limitations of the two most popular techniques (Kroenke & Gray, 2006).
Kroenke and Gray begin their article with a discussion on the role of data modeling in the development of information systems. The authors stress that the data model’s primary purpose is to capture “the user’s view of their [the user’s] world,” rather than to aid in database design (Kroenke & Gray, 2006). Having established the role of the data model, Kroenke and Gray proceed to list their minimum requirements for a data model:
- Sufficiently robust to readily express the users’ perceptions
- As simple as possible
- Independent of any physical database model
- Utilize domains with inheritable properties
- Readily support database migration
(Kroenke & Gray, 2006)
From a scholarly standpoint, Kroenke and Gray’s requirements are developed using logical arguments, but are not supported by any references to research literature. To demonstrate the viability of the minimum data model requirements, the authors present a simplified model that they refer to as a, “form based skeleton,” data model (Kroenke & Gray, 2006). After presenting the minimum requirements and their own simplified data model, Kroenke and Gray proceed to compare and contrast E-R and UML diagrams to the requirements. Figure 1 Kroenke and Gray Comparison of E-R and UML on Data Model Criteria shows the authors evaluations,
Figure 1 Kroenke and Gray Comparison of E-R and UML on Data Model Criteria (Kroenke & Gray, 2006)
Clearly, Kroenke and Gray’s evaluation means that there is a significant need for improved data modeling tools. Neither E-R nor UML meet or exceed data modeling requirements for any of the criteria. (Kroenke & Gray, 2006)
Data modeling is a critical activity for information systems design and development. Kroenke and Gray illustrate an important gap for information systems analysts: a lack of sufficient data modeling methodologies (Kroenke & Gray, 2006). The two most common data modeling tools, E-R and UML, are too complex and difficult, are too closely tied to database design, and fail to support group and domain attributes (Kroenke & Gray, 2006).
References
Hoffer, J. A., George, J. F., & Valacich, J. S. (2011). Modern systems analysis and design (Sixth ed.). Upper Saddle River, NJ: Prentice Hall.
Kroenke, D. M., & Gray, C. D. (2006). Toward a next generation data modeling facility: Neither the entity-relationship model nor UML meet the need. Journal of Information Systems Education, 17(1), 29. Retrieved from http://proquest.umi.com.library.capella.edu/pqdweb?did=1022838751&Fmt=7&clientId=62763&RQT=309&VName=PQD
