Normalizing the Database
Go Up to Establishing Relationships between Objects
After your tables, columns, and keys are defined, look at the design as a whole and analyze it using normalization guidelines in order to find logical errors. As mentioned in the overview, normalization involves breaking down larger tables into smaller ones in order to group data together that is naturally related.
Note
A detailed explanation of the normal forms are out of the scope of this document. There are many excellent books on the subject on the market.
When a database is designed using proper normalization methods, data related to other data does not need to be stored in more than one place—if the relationship is properly specified. The advantages of storing the data in one place are:
- The data is easier to update or delete.
- When each data item is stored in one location and accessed by reference, the possibility for error due to the existence of duplicates is reduced.
- Because the data is stored only once, the possibility for introducing inconsistent data is reduced.
In general, the normalization process includes::
- Eliminating repeating groups.
- Removing partially-dependent columns.
- Removing transitively-dependent columns.
An explanation of each step follows.