Sql Excel : What Is a Data Model?

The definition of the tables, the columns, and the relationships among them
constitute the data model for the database. A well-designed database actually has
two data models. The logical data model explains the database in terms that
business users understand. The logical data model communicates the contents of
the database because it defines many business terms and how they are stored in
the database.
The physical data model explains how the database is actually implemented. In
many cases, the physical data model is identical to or very similar to the logical
data model. That is, every entity in the logical data model corresponds to a table in
the database; every attribute corresponds to a column. This is true for the datasets
used in this book.
On the other hand, the logical and physical data models can differ. For instance, in
more complicated databases, certain performance issues might drive physical
database design. A single entity might have rows split into several tables to
improve performance, enhance security, enable backup-restore functionality, or
facilitate database replication. Multiple similar entities might be combined into a
single table, especially when they have many attributes in common. Or, a single
entity could have different columns in different tables, with the most commonly
used columns in one table and less commonly used ones in another table (this is
called vertical partitioning, which some databases support directly without
having to resort to multiple tables). Often these differences are masked through
the use of views and other database constructs.
The logical model is quite important for analytic purposes because it provides an
understanding of the data from the business perspective. However, queries
actually run on the database represented by the physical model, so it is convenient
that the logical and physical structures are often quite similar.