Semantic Modeling for Data

Chapter 5. Semantic Model Development

“If you mine the data hard enough, you can also find messages from God.” [Dogbert]

Scott Adams, Dilbert

So far we’ve seen what a semantic model can consist of, what phenomena should concern us during its development and usage, and what quality dimensions we need to evaluate before we use it. In this chapter, we’ll focus on the development process of a semantic model, looking at the challenges, steps, and activities it involves, and the methodological and technological support that is available for each of them.

If you are mainly a consumer of semantic models rather than a creator, this chapter will help you understand how challenging the development process can be, and how you can help model creators build you the models you actually need.

Development Activities

There are several methodologies and life cycles in the semantic modeling literature that attempt to define the steps and activities involved in the development of a semantic data model [93]. Some prescribe only a few high-level activities, while others contain more detailed and specific tasks. Some are specific to particular types of models and/or modeling frameworks, while others are more generic. Some are based on a more centralized philosophy, while others are based on a more collaborative one. And some enjoy comprehensive technology and tooling support, while others do not.

In my career as a semantic modeler I have had the chance to apply several of these methodologies in (nonacademic) ...