Date Added: May 2011
The "Normalization" is a practice used to design the relation(s) for a good database eliminating undesirable functional dependencies amongst that exist amongst attributes of the relation. The complexities involved in the normalization of relations, have mowed down vendors from automating the normalization processes. Although the keyword normalization is existing in the data manipulation language of Structured Query Language (SQL) standard This paper unravels the complexities involved in the normalization process and proposes an automatic methodology for refining the relations with normalization. The primary key for each relation is designed based on the superset of minimum attribute(s), which uniquely determines other attribute values of the tuple in the relation.