DATABASE MANAGEMENT SYSTEM (DBMS)

DATABASE MANAGEMENT SYSTEM (DBMS)

3.2.1 Data and Information

Data is the raw material from which information is derived as the end product.

Data represents a set of characters that have no meaning on their own, i.e., it consists

of just symbols. On processing, meaning is attached to data, which transforms into

information.

To illustrate the difference between these two terms let us consider an example. The

digits 050643 as such have no meaning. But if we are told that the first two digits

represented a month, the next two digits, a day of the month and the last two, a year, then

the set 050643 may represent the date of birth of a person. Processed in another manner

the same digits written as 643050 may represent the telephone number of an individual.

3.2.2 Database and Database Management System (DBMS)

Database

A database is a collection of logically related data arranged in a structured form designed

to meet the information requirements of multiple users. It may also be defined as a

collection of non-redundant operational data sharable between different application

systems.

Database Management System

It is a collection of software that is used to store, delete, modify and retrieve data that

is stored in a database. DBMS acts as an interface between the user and the data

3.3.2 The Goals of DBMS

The database management systems have the following goals :

i) To provide retrieval flexibility. It should be relatively easy to link data from different

files.

ii) To facilitate reduction of data duplication and elimination of multiple copies of a

master file. Data redundancy control helps in overcoming updating problems and

promotes data integrity.

iii) To ensure high level of data independence. The data is hidden from the programming

language, operating system and processing environment. It should be up to DBMS

to convert the stored data into a form that could be used in whatever language the

programmer desires to use.

3.3.1 The Need for DBMS

Let us consider the scenario of data processing that existed before the advent of database

management systems. In the file-oriented system, which was used at that time, a master

file (the file which contains all the up-to-date data on a subject) is created using a

programming language. Access techniques based on the requisite queries on the data are

embedded in the file at the time of its creation. Any change in the master file i.e., addition

of a new field or change in the structure of an existing field has to be implemented by

creating the file all over again along with the modified access techniques.

To illustrate the limitations of data management using master files, let us consider the

example of a database of an educational institution running professional courses. Let us

assume that the database consists of three master files of student, faculty and course

records. The student master file has been created using FORTRAN and has such fields

as student identification number, name, address, gender, course, high-school grade and

examinations cleared.

The faculty master file uses COBOL and consists of fields like faculty identification

number, name, gender, department, salary, qualifications and teaching hours. The course

master file is based on PASCAL and covers such data as course identification number,

course title, class number, section number and students attending the course.

Now, suppose there is a query to provide the names of all the female students being

taught by female faculty members. This query cannot be answered by the available

master files despite the fact that the data needed for the query exists in the database.

The difficulty lies in the fact that the needed data is available in two master files created

in different programming languages and having their own access techniques. To answer

the query a new master file with data items derived from the student and faculty master

files will have to be created and new programmes for accessing the data written. This

makes data retrieval cumbersome and time-consuming.

Take another situation when the accounts department of the institution in the example

also wants to use the database and needs the student master file with additional fields like

stipend paid, fees due, penalties charged, etc. To meet these requirements, another copy

of the student master file with new fields is created. Similarly, there may be copies of

faculty and course master files created to meet specific requirements. This results in

duplication of data i.e., data redundancy. In such circumstances it becomes difficult to

keep the master files identically updated i.e., propagate the updates in all the copies.

These limitation and drawbacks were at the core of development of database management

systems.

3.4.2 Functions and Components of a DBMS

Basically, there are only two operations that can be performed on data viz., retrieval and

maintenance. Retrieval refers to reading data from files to serve the information

requirements of a user and forms the most important function of a database management

system. Maintenance concerns changing of data in stored files.

Data maintenance involves three operations: addition, deletion and modification which

correspond to adding new records, deleting existing records and modify/updating values

in the existing records.

A database management system has two essential components: data definition part and

data manipulation part. Data definition part provides definition or description of database

objects and is written using data definition language (DDL). This part creates logical

structures of entities when a database is set up.