EA Information Model

EA Information Model Foundations

In the context of enterprise architecture, an information model provides a conceptual model of the enterprise architecture elements, their attributes and their relations to each other [RZM14; 11a; Buc+09c]. An overview of possible information model elements is provided by the TOGAF Architecture Content Framework [11b]. Figure 1 illustrates an example of an enterprise architecture information model which was created in accordance with the TOGAF Framework.

Even though many different information models have been introduced by literature and tool vendors [Bus+12], no overarching information model has been established in the field of EAM [BMS11a]. Therefore, some researchers claim that such models are enterprise-specific design artifacts [Buc+09d].

Advantages and Challenges of an EA Information Model

A good information model allows to structure elements of an enterprise architecture in a consistent way, which reduces the complexity of a system [11b]. Additionally, information about the system can be stored systematically, e.g. in a repository [Buc+09b], which enables advanced query and analysis opportunities [RZM14; 11b]. The results of such queries can be used as a basis for informed decision making [11b].

A main challenge when establishing an information model is to choose the best level of abstraction. On the one hand, if an information model collects too elaborate descriptions of the provided interfaces of the EA elements, it may lose its focus. Further, the data quality will decrease, because keeping all information up-to-date is connected with high efforts. On the other hand, the collected information needs to be detailed enough to provide a useful basis for decision making [Wit+07].

Approaches to Developing an EA Information Model

As indicated in Figure 2, there are two approaches to obtain an information model. On the one hand, an information model can be gained bottom-up [Mat11b]. This means, the information model of a system is gained by extracting it from existing information about a system [Mat11b]. One advantages of this approach is, that a system can grow freely and without having any structure forced on it right away [Mat11b]. Naturally emerging structures, processes and roles can be extracted and captured by the information model as soon as they surface [Mat11b]. Once captured, these commonalities of elements can be enforced to foster consistency of new EA elements [Mat11b]. Nevertheless, emerging information models are not always of high quality.