|4.25 Basic Method "Normalization" (NORM)|
The objective of method Normalization (NORM) is to generate data structures (entity types with attributes) so certain regularities or normalization rules are followed which might have the following effect:
A normalization of not normalized data structures (entity types with attributes) initiates a step-by-step splitting-up of these data structures and results in simpler normalized data structures, so-called normal forms.
It is essentially differentiated between five normal forms. However, only the first three of these normal forms have a practical importance. Each "higher" normal form fulfills the regularities of the "lower" normal form.
At first, the present complete entity-relationship models (E/R models) always have to be normalized up to the third normal form. In exceptional cases, a denormalization might be advisable when changing to the physical storage based on performance reasons. Together with LOGM - Logical DB Modeling and DNAV - Data Navigation Modeling, the method covers that part in subproduct SW Design (Database).Schema Definition which refers to the database management system (DBMS) oriented data structure.
The complete schema definition also requires the specification of external keys and indices as well as integrity conditions and possibly views. Furthermore, DBMS-specific storage parameters and all data elements must be completely defined.
|No.||Interface||Observation||Information in Annex 1|
|5.1||NORM-ER||Within the scope of the ER - E/R Modeling, the normalization is applied to change an E/R model comprising attributes into a normalized form.||4.12 Interface ER-NORM|
|GDPA Online Last Updated 01.Jan.2002 Updated by Webmaster Last Revised 01.Jan.2002 Revised by Webmaster|