The application of directed graphs to data modeling and normalization

As the ratio of personal computers (PCs) to mini computers and mainframes grows, more and more data processing applications will be run on PCs. Many of these applications will be designed using data and transaction modeling tools and normalization software. Data modeling and normalization are techniques not uncommon in many larger business enterprises today, but it is rare to find designers and programmers using data modeling tools for microcomputer applications. This is due to the lack of awareness of the benefits of data modeling and normalization and to the unavailability of automated tools for assisting the data base developer in the microcomputer environment.

This thesis describes an accompanying microcomputer software product, named the Data Modeling and Normalization System that provides such a tool for storing and updating data model user views, normalizing them, and designing a relational data base from the normalized scheme. The Data Modeling and Normalization System's data dictionary, data dependencies, and relational data base design data are stored in popular dBASE III data file format.

The normalization component of the Data Modeling and Normalization System is written in C and uses a directed graph data structure to represent the semantic data models used in the normalization process. Each step in the normalization process is described in Chapter VII. Such an in-depth description of how an algorithm for computerized normalization works has previously been unavailable because normalization software systems currently available on mainframe computers are proprietary.