Conceptual and Logical Design of Relational Databases

Publisher Summary This chapter gives an overview of the system-independent steps involved in the design of a relational database. The chapter presents the requirements analysis, the conceptual modeling, and the logical design as the main phases of database creation. During requirements analysis, the database designer specifies the requirements either for the whole organization or for each individual user group. To achieve a satisfactory design, a requirements statement should contain both, data and functional requirements. During conceptual modeling, the requirements are expressed in a so-called high-level data model. The chapter describes the original entity-relationship (ER) model, an extended ER model, and the Nijssen's information analysis methodology (NIAM) approach for conceptual design. The chapter presents a single example to summarize the essential features, constraints, and modeling constructs of these approaches. The view integration process is also described. To implement a conceptual schema in a database management system (DBMS), the schema has to be transformed in the specific data model supported by the DBMS in use. This transformation process is called “logical design,” and the chapter discusses the logical design of relational databases. These three main design activities are completely system independent. The final design step is called the “implementation phase” and consists of taking the results of the previous steps and actually defining the database using the data description language (DDL) of the DBMS in use.