A survey of primary and secondary keys through a case study

Two fundamental decisions face the designer of a database under an inverted file data base management system: (1) potential primary/secondary key definition and (2) selection among candidate keys for indexing. Functional dependencies and normalization have been defined as ways to construct potential primary keys. This tutorial paper suggests a set of both primary and secondary key types which has been helpful in developing potential keys. This set and the motivation for it are illustrated from an actual business database application. This application is further used to overview the benefits of a database design tool called ASSIST which evaluates alternative, potential keys.