1. Databases and their applications

A well-accepted definition of a database is a collection of interrelated data stored in a computer system. This definition is too general, but it can be made more specific when one considers that the data has some meaning for a specific organization or enterprise, that there will be a diverse group of users accessing the data, and that the database is built to satisfy a particular set of information needs within the organization.

In today’s database systems the data and the definition of the data, also called metadata (as described in Chapters 4 & 5), coexist to provide high-level access to the users of these data while assuring that the data is well kept and well secured. An application is a software system which accesses the database to satisfy some users´ information needs. Application programs and queries contain data access commands which are attended by a very complex software, the database management system software (DBMS).

The process of building a database system involves the design and implementation of  the database using a DBMS and the design and implementation of the applications that will use the database. The main activity of database design is to model the data in an abstract representation. The model is then implemented in the DBMS to be used. The representation of a specific database in the data model used by a DBMS is called the database schema. The schema holds the definition of the data and is written in the data definition language (DDL) of the DBMS used. The schema is compiled to generate the empty structures where the actual data will be loaded. Embedded in the schema are some instructions for enforcing integrity constraints. These constraints guarantee that the data is correct with respect to the domain. Additional constraints can be specified in several manners and must also be enforced.

After the data is loaded, the database is ready to be used. The users access the database through applications or directly by issuing commands in the now standard database language SQL (Structured Query Language). This way, the data and the programs that use it are kept separate. The DBMS allows access to the data and enforces controls to keep it correct and secure. The applications can then concentrate on specifying which data they need for their processing and not worry about how the data is obtained.

The advantages of following the database approach are manifold. This approach allows data independence, i.e.changes to the physical structures where the data is stored do not affect the operative applications; increases data sharing; facilitates the establishment of standards; reduces programming time by leaving most of the data validation to the DBMS.

In 40 years of database technology history and after the successful implementation of the relational model, there have not been many changes in the way the data is structured or in the data models used. During the 60´s and 70´s, the hierarchical and network models were prevalent. In the 80´s, the relational model became competitive and started taking over the market. Currently, there have been some penetration of the object-oriented model, although it still has a very small portion of the market. But there have been dramatic changes and rapid evolution in the interfaces to a database and in the heterogeneity and needs of the applications which use it. Web browser interfaces are very common today and very convenient. They are platform independent and allow simple and uniform access to databases. With the evolution of intranets and web browsers it is natural to integrate all the data and services of a business through a web site. We talk about this integration in the next section.

In the early days, a database was built for a single application. Today´s technology and practices exploit the database approach by building several applications around one, normally large and complex, database. The nature of these applications has also changed: what used to be a single application before, might be part of a more complex and sophisticated application today. For example, a human resources system includes recruitment, promotions, payroll, training and development, and termination. Each of these components could have been an application before and is now part of a large system. Additionally, with an integrated view of an enterprise, all the applications must interact and their data must be combined and analyzed to provide consolidated information to the strategic levels of the organization.

The database contains the data and the rules which define data correctness. The applications implement the dynamics of how the data is updated and used to fulfill information and knowledge requirements.