Data Modeling Helps Build Successful DatabasesWant to make sure your database contains all the information needed to create the reports and give you the information you need? Don't skip the data modeling step. For those who build databases, this data model is like an architect's plans. Just as a builder would be foolish to start a building with no clear plans, building a database with no clear plan can prove foolhardy also. Since data modeling is the most time-consuming and labor-intensive part of building a database, some would just prefer to skip it altogether. If you want your database to perform the functions you need, this is a bad idea. Just as you would begin to find problems right away with a house built with no plans, you would also find that without a data model, your database will also have problems. As these problems become obvious, you will probably spend more time trying to correct your poor planning than you would have by creating a proper data model in the beginning using a data modeling tool. When conducting the process of data modeling, workers usually create a data model working with data modeling software. Data modeling tells a computer or network how it should use and organize data. Commonly used types of models include hierarchical models, network models, relational models, object-relational models and object models. Other less common types of models include associative models, concept-oriented models, multidimensional models, star schema models and XML database models. Without data modeling, it is nearly impossible for two computer systems to share data. This is because one computer system may use one set of data written in a particular language while the second computer may not be able to read it. If you apply a data model to both systems, however, they will both be able to understand the data that presented to them. Data modeling is the planning process that must occur before you can build the data model on the system. Before you can plan your data modeling however, you must identify what data each system will share. If your systems are sharing tables of data with structure, such as those used in Structured Query Language (SQL), then you have to think about table structure as well as how you should link those tables. Other systems, such as mainframes use a whole different dataset and programming languages so connectivity between them is different. This changes the way you will plan your data modeling. PC systems connecting to mainframes may cause even further planning because the programming languages and datasets are sometimes so much different. The point is that data modeling is an important step in any IT manager's job. If the IT manager fails to plan how his system is going to work, it probably won't perform in the way he wants. |