Semantic Modeling for Data

Chapter 12. Representation Dilemmas

Design is the method of putting form and content together. Design, just as art, has multiple definitions, there is no single definition. Design can be art. Design can be aesthetics. Design is so simple, that’s why it is so complicated.

Paul Rand

When developing a semantic data model, it is often the case that a particular piece of information can be represented in more than one legitimate way, even within the same representation language (e.g., deciding whether an entity should be represented as a class or an individual). Each way has different strengths and weaknesses that you need to be aware of, both as a creator and a user of a semantic model. As a creator because you will need to pick and implement the way that is best for your case, and as a user because it will help you select between semantically equivalent models that don’t have exactly the same capabilities.

This chapter covers some common representation dilemmas, with particular focus on when and how you should represent vague elements by fuzzifying them.

Class or Individual?

In Chapter 2 we saw that, in some cases, the modeling language you use obliges you to decide whether an entity should be modeled as a class or an individual (i.e., as an entity that has no instances). We also saw that one problem with having to do such a selection is that there are several entities that can be legitimately modeled as both a class and an individual (e.g., Eagle and Data Scientist).

A second ...