TABLE OF CONTENTS Data normalization is a process by which the contents of the database (i.e., data) are made internally consistent. While much of the normalization can be done using data input validation (see next chapter, Data Validation, Editing, and Quality Control), this is not always possible when data are imported from external files, data tables, or legacy databases. However, the process is essential for successful data extraction, manipulation, reporting, and utilization.
This chapter defines the basic process of data normalization, explains why the process is so important, and provides examples of problems that can arise when normalization is done improperly. Several examples of normalization methods are demonstrated and compared.
Database normalization refers to the internal structure and makeup of data tables and the links between them. Database normalization is a complex issue, but one that is generally addressed by the DBA. The basic goal of database normalization is to ensure that key data elements are maintained in a regular, organized manner with no repetition or unnecessary duplication of information from table to table within the database.
The concepts of data normalization are not as well defined as formal database normalization. The DBA ensures that tables are fully normalized and that the system is tuned; responsibilities rarely extend to the issue of data accuracy. A successful and useful database is one where the data are internally consistent and accurate. Internal inconsistencies and errors generally bring into question the accuracy of the entire database. A single major...
