Excerpt from a new blog focused on modeling realities
Let the debunking begin. Last time we said that data modeling is difficult for various reasons, some legitimate and some silly, and that we’d scrutinize some of the silly ones. One reason: People misunderstand the nature of conceptual modeling.
First some basic definitions. […]
- A conceptual data model is a technology-neutral expression of the kinds of data that some person or organization values. To stress the point, I sometimes say “technology-agnostic” or even “technology-apathetic.” When you are building a conceptual data model, you should not care about technology. (Hard to realize in practice, but let’s at least get the definitions right.)
- A logical data model is also an expression of kinds of data, as that data would be represented by a particular technology paradigm such as the relational model.
- A physical data model is also an expression of kinds of data, as that data would be represented in a particular implementation in a particular product, such as a specific version of Oracle Database or Microsoft SQL Server.