b. data modeling
Data modeling is the analysis of data objects that are used in a business or other context and the identification of the relationships among these data objects. Data modeling is a first step in doing object-oriented programming. As a result of data modeling, you can then define the classes that provide the templates for program objects.
A simple approach to creating a data model that allows you to visualize the model is to draw a square (or any other symbol) to represent each individual data item that you know about (for example, a product or a product price) and then to express relationships between each of these data items with words such as "is part of" or "is used by" or "uses" and so forth. From such a total description, you can create a set of classes and subclasses that define all the general relationships. These then become the templates for objects that, when executed as a program, handle the variables of new transactions and other activities in a way that effectively represents the real world.
Several differing approaches or methodologies to data modeling and its notation have recently been combined into the Unified Modeling Language (UML), which is expected to become a standard modeling language.
For more information:
Best Web Links: Modeling
White paper: Modeling the data warehouse and data mart
White paper: Dimensional data modeling
Book excerpt: Introduction to Oracle Design
Do you have an idea for a quiz question topic? Let us know!
This was first published in August 2003