Database Structures

In database management a structure is the plan or method by which data is logically organized in a system. A database structure is also called a schema. A schema or structure includes a data model which precisely describes the basic data elements or attributes of a field for example how many characters are contained in a field. The schema also assigns identifiers for referencing data.

To choose a database structure or schema one must assess the needs of the organization. This is determined if the database models will be used locally for local processing and local users only or the data model will be used in a distributed data base system for all involved parties of the system. Information limited to local use only is called “Restricted or Local Data Elements” while information that is available for a distributed database system is called “Global Data Elements”.

Database schema allows records to be located, edited, and deleted and any related activities. This means that schema ensure that a whole database does not have to be redone every time changes are to be made. A schema enables one to make changes to the specific area of the database and not the whole database hence save a lot of effort. Here are the four most common database schemas.

List Structure
List structure enables records that are out-of-sequence to be treated as if they were in a logical sequence. This is made possible by use of numeric pointers that can take next records as if they were in a sequence. Data needs to be stored serially on a direct access device for random locating.

For example data may not be stored alphabetically in the database but can be retrieved in an alphabetical order. How? The data is randomly stored but has a linkage address or pointer which is automatically assigned by the system during storage. The computer uses this address to access the names and order them alphabetically and then pass them to the interface for the user to sort and retrieve.

Tree or Hierarchal Structure
Tree structure has two types of records that enable its implementation. These are the master record and subordinate records. Master records are sometimes called parent or owner records while subordinate records are called children records. Subordinate records have only one master records linked to them in this structure while the master records may have more than one subordinate record.

The computer first finds the master records and then branches down to subordinates further till it obtains the required data. This is why it is called hierarchal structure. Tree or hierarchal structure is therefore ideal for data that are inter-related.

Network Structure
Network structure like tree structure has master and subordinate records. But in network structure, a master record may have more than one subordinate record and a subordinate record can have more than one master record. This is what differentiates the two structures.
One master record in network structure is thus able to direct a computer to more than one subordinate record.

Relational Structure
Relational structure uses tables that have relationships to search data. It does not need any pointers to locate data. The elements that will be searched in a database may be constructed after the database has been constructed.
Relational structure takes lists from one table and uses its relationships to other tables to construct the needed search results. This means that data does not have to be in one file because the computer is able to branch from one file to another. This makes the relational structure the most flexible schema in database management.