Representation or Implementation Data Model
The Conceptual Data Model is transformed into the Representation Data Model. Representation data model uses concepts that are understood by the end-user and are also close to the way the data is organized in the computer. These models hide the details of data storage. The data models are broadly classified as traditional data models that include—(1) hierarchical, (2) relational, and (3) network data models. Object-relational data model is an emerging data model.
Relational Database Model
The Relational Database Model was proposed in 1970 by E. F. Codd. The first commercial system based on the relational model became available in early eighties. Relational database model is the most common type of database model. Table, record, field, key, and data values are the terms associated with a relational model. The data elements are stored in different tables made up of rows and columns. The data in different tables are related through the use of common data elements. We briefly define the terms as follows—
Table “Employee”
· Data Values Data values are the raw data represented in numeric, character, alphanumeric, or alphabetic form. Examples of data values are ‘Abhinav Bindra’, ‘26’ ‘shooting’, “Chandigarh” etc.
· Field or Column Data values or data is stored in a database as fields. For an item or object, a field holds the information of it. For example, individual fields or columns are name, age, address, hobby etc. for an object ‘student profile’.
· Record or Row A group of data for related field is called a record. For example, for object ‘student profile’, the related fields are (name, age, address, and hobby). The data about the student #1 (‘Abhinav Bindra’, 26, ‘Chandigarh’, and ‘Shooting’) is one record.
· Table A collection of logically related records form a table. A table for an object has rows and columns. The table is organized as a set of columns, and can have any number of rows. For example a table for student profile can have four columns namely, name, age, address, and hobby, and have records or rows having data of 30 students.
· Key A key is an identifier in a table that uniquely identifies a row in a table. The key identifier can be the value of a single column or of multiple columns. A key is generally also referred to as the primary key of the table. The primary key is a unique identifier for the table. The column or combinations of columns that form the primary key have unique values. At any time, no two rows in the table can have same values for the primary key. For example, in a student table, each student has a unique student_rollno, which forms the primary key.
In its simplest form, a database may contain a single table stored as a file. In more complex databases, a collection of related tables may be stored together. For example, a student database may contain a table for student profile, student result, student courses etc. Figure shows a table “Employee”, with fields, records, data values & primary key. Several commercial products like DB2, ORACLE, SQL Server, SYBASE, and INFORMIX are relational databases. The Relational Data Model is extended to include the object database concepts thus forming the Object-Relational Data Model.
Hierarchical Database Model
The Hierarchical Database Model was developed by IBM and is the oldest database model. The hierarchical database model is defined as follows—
· The schema of a hierarchical database is represented using a Sree-structure diagram
· The nodes of the tree represent a record type. A line connecting two nodes represents the link.
· The schema is based on parent-child relationship.
· A parent can have none, one, or more children. A child can have only one parent.
· The parent-child relationship is suited for the modeling of one-to-many relationship between two entities.
· It is difficult to implement a many-to-many relationship using hierarchical database model.
· Some of the hierarchical database implementations are the IMS system from IBM and System 2000 from MRI systems.
Hierarchical database model
Network Database Model
The Network Database Model was formalized by Database Task Group (DBTG) group of Conference on Data Systems Languages (CODASYL) in the late 1960s. The Network Database Model is defined as follows:
· The schema of the network database model is represented using a data-structure diagram.
· The boxes represent the record type and the lines represent the links.
· The schema is based on owner-member relationship.
· The entity type is represented using record type and relationship between entities is represented using set type.
· A set type can have more than one record type as a member but only record type is allowed to be the owner in a set type.
· The owner-member relationship is suited to represent one-to-many relationships. The one-to- many relationship is converted into a set of one-to-one relationships.
· The network model handles many-to-many relationship by converting it into two or more one- to-many relationships.
· Some of the network database implementations are IDMS, DMS 1100 and IMAGE.
Network database model
The hierarchical model is simple to construct. However, the network model is able to handle complex relationships easily. Both hierarchical and network models are also called Segacy models as most current systems support the Relational Model.
Low-Level or Physical Data Model
The Physical Data Model describes the internal storage structures, the access mechanism, and the organization of database files. Physical level describes how a record (e.g., customer) is stored. Sequential files, direct files, and indexed sequential files are the different types of file organizations at the physical level.