ABSTRACT
Database systems have a number of advantages over using simple files or spreadsheets to store and manage information. However, simply using a relational database management system to store data does not guarantee advantages or improved efficiency. Users developing new systems or migrating data from spreadsheets may simply create tables and databases based on how they are used to seeing or using data. As a result, the resulting database may contain the same data in multiple locations, making it difficult to understand and use information, and normally guarantees inconsistencies in the values stored.
Database design begins with the process of normalization, where the design team creates a set of entities and attributes in a logical data model that describes the user’s data requirements. The data model is then reviewed and modified by a rigorous set of rules to take it to, respectively, first, second, and finally third normal form, where data elements are logically grouped where they belong. These rules at first appear to be very rigid and restrictive but with practice become second nature to team members and they learn to think in third normal form when defining data requirements.
