Classification of database management system is based on various parameters such as the kind of data model used to construct the DBMS, the number of users that will be using the database system, the way in which the database is distributed. In the section ahead, we will be discussing all the criteria on which the database system can be classified.
Classification of Database Management System
- Based on Data Model
- Based on Number of Users
- Based on Database Distribution
- Based on Cost of Database
- Based on Usage
- Based on Flow Control
1. Based on Data Model
The data model defines the physical and logical structure of a database which involves the data types, the relationship among the data, constraints applied on the data and even the basic operations specifying retrieval and updation of data in the database. Depending upon how the data is structured, data models are further classified into:
a. Relational Data Model
In the relational data model, we use tables to represent data and the relationship among that data. Each of the tables in the relational data model has a unique name. A table has multiple columns where each column name is unique. A table holds records which has value for each column of the table.
We also refer to the relational database model as a record-based data model as it holds records of fixed-format. The relational database model is the most currently used data model.
b. Entity-Relationship Model
The Entity-Relationship model (E-R data model) represents data using objects and the relationship among these objects. These objects are referred to as entities that represent the real ‘thing’ or ‘object’ in the real world.
Each entity in the E-R model is distinguishable from other entities in the model. Like, relational model, the E-R model is also used widely used to design the database.
c. Object-Based Data Model
Nowadays, object-oriented programming such as Java, C++, etc. is widely used to develop most of the software. This motivated the development of an object-based data model. The object-based data model is an extension of the E-R model which also include notion for encapsulation, methods. There is also an object-relational data model which is a combination of the object-oriented data model and relational data model.
d. Semistructured Data Model
The semistructured data model is different from what we have studied above. In the semistructured data model, the data items or objects of the same kind might have a different set of attributes. The Extensible Markup Language represents the semistructured data.
Before the above data model was introduced the data models like hierarchical data model and network data model were used to design a database that is still used in some legacy applications. The hierarchical data model stores the data in the form of records and uses a tree structure to represent these records. The record is arranged in a tree structure where there is a single parent record for each child record.
2. Classification Based on Number of Users
The database management system can also be classified on the basis of its user. So, a DBMS can either be used by a single user or it can be used by multiple users. The database system that can be used by a single user at a time is referred to as a single-user system and the database system that can be used by multiple users at a time is referred to as a multiple user system.
3. Based on Database Distribution
Depending on the distribution of the database over numerous sites we can classify the database as:
a. Centralized DBMS
In the centralized DBMS, the entire database is stored in a single computer site. Though the centralized database support multiple users still the DBMS software and the data both are stores on a single computer site.
b. Distributed DBMS
In the distributed DBMS (DDBMS) the database and the DBMS software are distributed over many computer sites. These computer sites are connected via a computer network. The DDBMS is further classified as homogeneous DDBMS and heterogeneous DDBMS.
- Homogeneous DDBMS: The homogeneous DDBMS has the same DBMS software at all the distributed sites.
- Heterogeneous DDBMS: The heterogeneous DDBMS has different DBMS software for different sites.
4. Based on Cost of Database
Well, it is quite difficult to classify the database on the basis of its cost as nowadays you can have free open source DBMS products such as MySQL and PostgreSQL. Although the personal version of RDBMS can cost up to $100.
The large systems along with the components that can handle distribution of database, replication of database, parallel processing, mobile capability and so on can be sold in the form of licenses. The site license allows unlimited use while another kind of license limits the number of concurrent licenses.
Some systems with single-user versions such as Microsoft Access are sold per copy or included in the configuration of your desktop or laptop.
5. Classification Based on Usage
On the basis of the access path that is used to store the files, the database can be classified as general-purpose DBMS and special-purpose DBMS. The special-purpose DBMS is the one that is designed for a specific application and it can not be used for another application without performing any major changes we refer to this as online transaction processing (OLTP). The OLTP system supports a large number of transactions concurrently without any delay.
The general-purpose DBMS is the one that is designed to meet the need of as many applications as possible.
6. Based on Flow Control
Based upon the flow of control from application to DBMS the database management system is broadly classified into two types active database management system and passive database management system. Let us discuss each in brief.
With the passive database management system, the user needs to specify the query to the current state of the database system to retrieve the desired information. It is similar to traditional DBMS where the user or the application program is responsible to initiate the operation. As the application forwards the request to DBMS and waits for DBMS to process the query or requested operation and provide a possible answer.
The requested operation may be used to define or update the schema of the database or retrieve or update the data within the database. The passive database management systems are also referred to as program driven systems.
The active database management system on other hand are referred to as data-driven systems or event-driven systems where the control flow between the application and DBMS is based on the occurrence of an event. Inactive DBMS the control flow is two ways i.e. application can call DBMS and even DBMS can call application.
Inactive DBMS the users need to specify what data they require. The DBMS processes the request and if the requested information is currently available then it provides it to the user. If the requested data is not available currently then inactive DBMS the scope of query also includes the future data i.e., the DBMS will monitor the arrival of desired information in future data and will provide it to the relevant user.
So, this is all about on what basis you can classify the database management system.