Question

1. In general, compare the Static data model, to the Dynamic data model, to the Normalized data model.

2. What does the Dynamic data model add? (behavior classes, communication diagrams)

3. What does the Normalized Data Model add or change (association classes, FK, attributes)?

Answer 1

1)

Most data models are static, in that they represent the properties of, and relationships between, business entities at a point in time. However, for a system to properly function over time, its data model must be designed to support data update in response to changes in the real world. Dynamic Data Modeling covers not only static data structures but update policies, by considering issues such as:

• what real-world changes must be captured in the database?
• what are the requirements for preserving a record of the historic state of the attributes and relationships of any entity?
• why must changes in attributes and changes in relationships be dealt with differently?
• do we also need to record changes in our state of knowledge of the real world?
• what aspects of the time dimension need to be taken into account?

There are two primary advantages of having a highly normalized data schema:

1. Increased consistency. Information is stored in one place and one place only, reducing the possibility of inconsistent data.
2. Easier object-to-data mapping. Highly-normalized data schemas in general are closer conceptually to object-oriented schemas because the object-oriented goals of promoting high cohesion and loose coupling between classes results in similar solutions (at least from a data point of view).

You typically want to have highly normalized operational data stores (ODSs) and data warehouses (DWs).

The primary disadvantage of normalization is slower reporting performance. You will want to have a denormalized schema to support reporting, particularly in data marts.