There are several types of database structures, such as flat-file, relational, hierarchical, network and object-oriented ones.
A database that consists of a single data table is called a flat-file (sequential file) database. Flat-file databases are useful for certain single-user or small-group situations, especially for maintaining address lists or inventories. Flat-file database systems are easy to learn and use, but difficult to maintain and limited in their power. When numerous files exist (one for each table or related document), there is often a lot of data redundancy, which increases the chance for errors, wastes time, and uses excess storage space.
A relational database is made up of a set of tables, and a common field existing in any two tables creates a relationship between the tables. For example, a Customer ID Number field in both the Customers table and the Orders table links the two tables, while a Product ID field links the Orders and Products tables. The relational database structure is widely used in today's business organizations. In a business, a typical relational database contains such data tables, as Customer information, Employee information, Vendor information, Order information, Inventory information.
The hierarchical database is an older style of database. The tables are organized into a fixed treelike structure, with each table storing one type of data.
The trunk table (the main table) stores general information. Any field in that table may reference another table that contains subdivisions of data. Each one of those tables may, in turn, reference other tables that store finer subdivisions of data. The relationship between tables is said to be a parent- child relationship, or one-to many relationship, with any child table relating to only one parent table. Each parent table may have many child tables, but each child has only one parent. Hierarchical
databases require little duplicated data and may locate data quickly. However, the tables' fixed relationships limit the flexibility of the database, making some kinds of queries or reports difficult or impossible.
The network database model is similar to the hierarchical structure except that any one table can relate to any number of other tables. The network database's tables, therefore, are said to have a many-to-many relationship. Like the hierarchical structure, the network database is used in older (primarily mainframe) systems.
The object-oriented database (OODB) developed in the late 1980s, groups data items into complex items called objects. These objects can represent anything: a product, an event, a customer complaint, or even a purchase. An object is defined by its characteristics (e.g. text, sound, graphics, video), attributes (e.g. color, size, style, quantity, price), and procedures (the processing associated with an object).