UNIT - IV

DESIGN PROCESS
The design process:
design of output, human factors in screen design,
issues involving human interaction, output layout
design. Design of input and control: data
capturing, input validation, design of input of
online systems, file and database development,
file organization, systems development in a
database environment, design in distributed
environment.
Introduction:
1. The design process:
Human interaction is an essential prerequisite in any
system. The interference of human makes the system work
or do anything. This involves the following steps:
1. Identify the specific output needed to meet requirements.
2. Select methods of presenting information.
3. Create document, report or other formats that contain
information.
1.1design of output
Output from a system accomplishes any of the following:
● Convey information about past activities, current statuses or projections
of the future.
● Signal important events.
● Trigger some action.
● Confirm an action.
An output design can be termed as good only if it meets the requirements
of the users.
The different types of outputs are
● A report
● A document
● A message
Based on the application the output may be printed or displayed.
The contents of the output may originate from:
● A data store
● From e process or system activity
● Directly from an input source.

The key questions to be answered during output design are:

● Who will receive the output? Are they inside or outside the
organization?
● What is the planned use of the output?
● How much detail is needed?
● How often is the information needed?
● By what method is the output to be presented? (display,
printed, audio output).
1.2Human factors in screen design
Mooer’s law: An information retrieval system will tend
not to be used whenever it is more troublesome for a
customer to have information than for him not to have
it.
Hansen’s user engineering principles:
● Minimize memorization
● Optimize operations
● Engineer for errors
The three levels where attention has to be paid to user
psychology
Functional level: This comes under the analysis phase. Deals
with questions like what functions should be performed by
the people and what by the machine. People should be able
to:
✔ Act flexible and intelligently when exceptional situations arise.
✔ Deal with ambiguities of data.
✔ Deal with errors caused by data entry.
Procedural level: This comes under the design phase. This
deals with the question of how to implement the above
functions. Factors to be considered are:
✔ Hardware: Type and location of terminals, type of keyboard etc.
✔ Language of communication.
✔ Conflicting user needs: cater to the needs of both experienced
and new users. & ✔ User acceptance
Syntactic level:
This comes under the implementation phase. It deals in the following
areas:

✔ System sensitivity-the following guidelines can be followed:

● Display a small amount of information at a time.
● Display one idea at a time.
● Always let the system respond to the user (either display the needed
information or display a message to indicate that processing is going
on).

✔ Display characteristics – the guidelines are:

● Make the instructions to the operator stand out.
● Clear up the screen whenever possible.
● Make help easily available.
✔ Availability factors – the guidelines
are:
● Make the system easily available to a
large number of users.
● Balance a trade off between a too
overloaded a system that no user can login
to a too restrictive system that the system is
not popular.
1.3Issues involving human interaction in
screen design:
The issues involved are
● Design consistency
● Dialogue design
● Performance criteria
● Bullet proofing
● Screen formatting software
Design consistency
Consistency in output reduces the learning time of the users. There should be
consistency in:
● Keying procedures
● Screen format
● Display area of prompts and messages
● Language used in prompts, messages and help information
Dialogue design
A prompt or message in a dialogue between the user and the system. There are other
types of dialogues also:
● Simple query
● Mnemonics
● English language query
● Function keys
● Building a record on the screen
● Operation instruction
● Menu selection
● Use of displayed format
● Text editing
● Hybrid techniques
Performance criteria:
● Response Time : Response time is defined as the
interval between the user’s pressing the last key in the
input operation and the terminal’s displaying or typing
the first character of response. The response time for
complex queries may be high in which case the system
must display a message asking the user to wait. There
should not be a too high standard deviation in
response time. The standard deviation should be less
than or equal to half of the mean.
Response time depends on:
✔ Speed of the computer
✔ Core storage available
✔ Operating System
✔ Number of users simultaneously online
✔ Efficiency of the program

● Form of system output: the output should:

✔ Be clear and legible
✔ Have prompts and messages in bold face
✔ Have a blinking cursor to attract user’s attention
✔ Have messages appearing in the same area of the screen

● Unit cost as performance measure: a long response time keeps a user waiting. An
economic value can be assigned known as the waiting cost. The efficiency ratio is also
sometimes used to measure system performance. It can be defined as:
ER= mean number of lines entered per unit time
But this ER may be misleading as the ratio on the experience of the user also.
Bulletproofing: Bulletproofing protects the
system from misuse of users and also avoids
confusing the user by misguided or imprecise
dialogue. Various steps for bullet proofing are:
● Non-automated checking-two people check the
design, one acts as the system and the other as the
user.
● Live operation with selected operators
● Dialogue control group
● Training users
Screen formatting software
Many software are available to design and
implement screens.
Some examples are:
● THE MENU SYSTEM
● CFI Operating Menu
● SCOPE
● FMS
● SCP22
1.4Design of output layout
An output layout is the arrangement of items on the output
medium. The layout should show the location and position of
the following:
All variable information
✔ Item details
✔ Summaries
✔ Control breaks
✔ Separators

All preprinted details

✔ Headings ✔ Document names and titles
✔ Corporate names and addresses ✔ Instructions
✔ Notes and comments
The layout form is tool used to design layouts. It tells the analysts exactly
where the data will be printed, how far away from the other details and
whether there will be enough space to include all essential details without
cluttering.
Conventions of data representation
● Variable information:
X – is used to represent alphanumeric data
9 – is used to represent numeric data
● Constant information: here the data must be represented on the
layout form in the same way if it is to be represented in the output,
Designing printed output:
Determine what items will appear on the printed report. The next step is
to determine the type and length of the data items from their data
dictionary entries. Use the layout screen to design the following:
Headings: these include
● Report titles
● Date and page number
● Column headings
Data and details: the data to be presented is represented using the X
and 9 conventions. Other editing characters are also
used(Z,B,$,ete,.).For example a eight digit alphanumeric data item can
be represented on the layout form as:
XXXXXXXX
(or)
X-------------X
It is necessary to define only once even though the data item may be
repeated. Eg:
ITEM CODE
XXXXXXXX
Summaries: the summaries, column totals, subtotals, etc must
also be defined in a similar way.

Must be designed to read from left to right and top to bottom.

● The most important items must be easiest to find
● All pages must contain the title, page number and date
● All columns should be labeled
● Abbreviations should be avoided
Designing of visual display output
The main differences here from printed output are:
● Less space available than on paper
● Instructions on how to use the display are required.
The display page is called as a screen or a panel. First the
characteristics of the screen are determined. These include:
● The physical dimensions of the screen
● Number of rows and columns of data that can be displayed
● Degree of resolution(high, medium or low)
● Number of colours
● Methods of highlighting
● Methods of intensity control
The screen is composed of different areas called
screen sections. Generally the top left section is
the most visible. This area must be used to
display the most important information. There
must different areas for:
● Headings and titles
● Contents of the display
● Messages and instructions
● Explanations
Multiple screen design: Sometimes more than one
screen is used to present information. For example one
screen can give general information and the next screen
can give more detailed information on any one item.
Windows design: Windows in output design are
subdivisions of the display screen that make it possible to
present different sets of output simultaneously. The output
sets can be from different programs or from the same
program. Windowing is used when it is necessary to:
● View different set of data simultaneously
● Switch between different programs
● Move information from one window to another
● Allow users to reposition information
Each single window must considered as a single screen
and must be designed with care using the above
guidelines for screen design.
Chapter2:
Introduction:
2.1Design of input and control:
Objectives of input design
● Controlling amount of input: an effective input design
should control the quality of input. There are two reason for
this:
✔ Reducing cost.
✔ Speed up the entire processing.
● Avoiding delay: A processing delay may occur from the
data preparation and data entry. This delay is called as a
bottleneck. A good input design will avoid processing delay
as much as possible.
●Avoiding errors in data:
The errors committed can be measured with
the help of error rate, which is the number of
errors made per unit time.
The error rate can be reduced by good input
design as the manner the data is entered
affects error rate. Another aspect of error
control is detecting errors as they occur.
Input design must also include validation so
as to check errors.
●Avoiding extra steps: As the volume of the transaction
increases even a small unnecessary step may add up to a lot
of wasted time. Therefore a good input design must avoid
any unnecessary steps.

●Keeping the process simple: Having excessive controls of

errors may lead to a difficulty in using the system. The users
may not accept, a complex or confusing input design as
easily as a simple one.
2.1Data capturing
Data capture guidelines:
The analyst must capture only those data that must actually be
input. There are two types of data that must be input:
● Variable data: those data items that change for each
transaction.
● Identification data: the element of data that uniquely identifies
the item being processed.
There are three types of data that must not be captured
● Constant data: data that are the same for every entry
● Details that the system can retrieve
● Details the system can calculate
Details of source document:
The source document is the form on which data are initially captured.

For the design of the source document it must be first decided on

what data must be first captured. Then the layout is developed.
Layout :
The layout organizes the document by placing important information
where it will be noticed and establishing a proper sequence of items.

The data that is needed must be requested in a logical sequence.

The document must be divided into different zones or information

groups. Inside each information group the data must be in a logical
order.
Captions and data capture
Captions on the document tell the users what
data to provide and where they should be
entered.
The captions should be brief, easily understood
with standard terms. Abbreviations should be
avoided. They are different types of captions.
⮚Line captions:
▪Before Line
▪After line
▪Above line
▪Below line
⮚Boxed captions:
●Inside box
●Below box
●Above box
• Coding methods
Coding reduces input, controls errors and speed up the
entire process. A code is a brief number, title or symbol
used instead of more lengthy descriptions. The different
types of codes are:
• Classification codes:
Classification codes place different entities in distinct
groups called classes. Here the user will classify the event
or item in one of the classification codes and write it in the
source document.
• Eg: when vehicles are classified into different classes in
a toll booth.
Function codes:
Function codes state the activities or work to be
performed without spelling out all the details in narrative
statements. This type of code is used in transaction
processing to specify the transaction. The function code
may be used to determine the contents of input record.

Sequence codes:
Sequence code is either numbers or letters assigned in
series. Sequence codes are used to
• Preserve the order in which the transactions occurred.
• Identification purposes
Significant –digit subset codes:
These codes can provide a lot of information .the
codes are divided into subsets or sub codes; characters that
are part of the identification number and that have special
meaning. Eg 05MSS10

Mnemonic codes:
Mnemonic codes use letters or symbols that
communicate visually. Eg:Tv-CL-21 describes a 21-inch
color television set.
Methods of data capture:
Source data capture with key punching:
The input steps are required are:
• write the data on the source document
• If necessary code the data into computer acceptable
form
• Key punch the data into cards
• verify the cards on a verifying machine
• Batch the cards and read them into the computer.
• Validate the data as they read into the computer.
• Process data.
Source data capture with key to storage:
The data are entered through a workstation, which
includes a keyboard and a display. The data are recorded
on a magnetic storage media.
The input steps are:
• Write the data on the source document.
• If necessary code the data into computer acceptable
form
• Enter the data through the keyboard.
• Read the data from storage and validate the data as they
are read into the computer.
• Process data.
Source data capture with scanner.
This involves optical character processing. The steps involved
are
1. Write the data on the source document
2. Batch the source documents together the read them
into the computer.
3. Validate the data as they are read into the computer.
4. Validate the data as they are read into the computer.
5. Process the data.
2.2Input validation:
There are three main categories of methods:
• Checking the transaction
• Checking the transaction data
• Changing the transaction data
• Checking the transaction:
This method checks the validity of the transaction. Transaction can
be valid because they are
• Out of order
• Incomplete
• Unauthorized
• Checking the transaction
Batch controls:
• In batch processing there is a delay between the occurrence of a
transaction and the time they are processed. The transactions are delayed
by accumulating the transaction into group called batches. The various
batches are
• Fixed batch size: here each batch contains a fixed number of
transactions. This helps to check if all the transaction in a batch is present.
• Numbering the batches: here a serial number is assigned to each batch.
The ensures no batch is over hooked or lost.
• Batch totals: some data item in each transaction in batch is added. This is
known as the batch total. This total is rechecked when the batches are
processed. This ensures that all transactions are processed properly.
 Transaction validation:
• Users try to process data in unexpected ways either intentionally or by
mistake. Validation procedures must be provided to take care of such
cases.
Sequence tests
• These tests use codes in data like serial numbers. There are two reasons
for their use:
• The order of transactions
• Missing item
Completeness tests
• These tests assure that the transaction data are accurate and complete.
Checking the transaction data
• Valid transactions may have invalid data. The checks that can be used
are:
Existence tests
Some data have to be present. For example in a sale the quantity purchased
must be present. These must be validated. For some items the data is
present depending on the type of the transaction.
Limit and range tests:
These tests check the reasonableness of the transaction. Limit tests check
either the minimum or maximum value, whereas the range tests check both
the minimum values and maximum values.
Combination tests:
These tests validate several data items jointly. The value of one element will
determine if the value of the other element is correct or not.
Duplicate processing:
In sensitive applications data will be processes more than once, either in
different ways or by different machines.
Modifying the transaction data:
Here the validation is done by modifying the transaction data. There are two
methods:
Automatic correction: This method minimizes the number of separate error
correction steps or rejection of transactions during processing. This way the
error is detected and corrected automatically. Eg: insertion of leading zeroes.
Check digits: There are two common errors when data are captured correctly
but entered incorrectly:
• Transcription errors: When one character is written as another.
• Transposition errors: when the position of the character is changed.
The check digit adds an additional digit to the data element. The digit added
is called as the check digit and can be calculated by some formula or method.
Whenever the data item is entered the check digit is calculated and seen if it
is the same. If there is a change then the data item has been entered
incorrectly.
2.3 Design of Online systems
Online systems are different from other systems in the following ways:
• They require immediate response to the user requests.
• Less predictable demand
• Direct contact between computer and user.
• Immediate response: They require from the system any one of the
responses immediately
• Acknowledgement
• Error message
• Signaling that processing has begun
• Requested information
Less predictable demand: Demands can occur at various times and at various
intervals and these demands must be dealt with immediately.
• Direct contact between system and user: There is no buffering of users
from the system. The following factors must be considered:
• Design of interface is very important.
• More security must be provided.
• Online systems are typically integrated with the operations they support.
Therefore their failure will stop the entire process and hence more
importance must be given to prevent system failure and data loss.
Interface: Boundary between the user and the
computer application.
Purpose of an interface:
Used to tell the system what actions to take
Facilitate the use of the system
Avoid user errors
Characteristics of an interface
Devices used to enter and receive data: Common
devices used in online systems are keyboard, mouse,
light pen, scanner, touch screen, voice etc,.
Dialogue: guides the interaction between the system and the user.
Methods and patterns used to display information: this is the
organization of output. How the physical area of display is structured
and how important data are highlighted.
Actions occurring at the interface
Navigation: navigation is moving from one place to another inside
the system.
Processing actions: the following are the important processing
actions:
Data entry
Data editing
Data storage
Data retrieval
Receiving messages
Messages are used to inform the users of:
When to take actions
Status of certain events and activities
When a task is completed
What errors of interruptions occur
Designing dialogue
Dialogue charts can be used to depict activities of an information system
in a graphical way.
They serve as road maps through the system. Short cuts will be useful to
quickly move from one screen to another.
Major dialogue design decisions
Overall dialogue strategy
Data entry dialogue
Paging and scrolling
Messages and comments
User navigation
Key assignment
Help systems
Dialogue strategy: the overall dialogue strategy can be one of the
following
Menu Driven
Keyword
Question/answer
Menu driven dialogue: a menu is an exhaustive list of available system functions for the user to
choose from. The menu alternatives can be chosen in different ways
• Keyboard
• Mouse
• Touch screen
• Light pen

Position of the menu: transaction-processing menus occupy entire screens. A single column or two
columns of menu items can be used. When business data has also to be retained, then the menu is
placed on the top or bottom of the screen or vertically on the side of the screen. Pull down menus can
also be used. Nested menus are generally used when:
• The number of alternatives is too large to fit on a single screen.
• A series of interrelated choices have to be made
• A complicated application may need each choice having progressively more detail
Keyword dialogue: in this strategy users invoke processing activities by entering a command the
system understands. There are two methods:
• Single command form
• Mnemonic command form
• Natural language form
Question/answer dialogue:
In this method the system presents a question to the user and the answer guides the resulting
processing. The answers can be both in a yes/no format and in narrative format.
Data entry dialogue:
Data entry template: a data entry template is a form
or an outline showing the information to be entered.
In addition to titles and headings it contains labels
that identify the data to be entered.
Question and answer prompt:
Users are prompted for data by questions the
system poses. The question is displayed only after
the present question has been answered.
Editing in online systems:
Editing refers to both making changes in already
entered records and also in deleting records. There
must be method for identifying the data to be edited.
Deleting records require the user to identify the record
to be deleted and instruct the system to be deleting the
record.
Paging and scrolling:
Paging and scrolling are used when there is too
much information to display on a single screen. When
using paging the user can move from one page to
another. A page number identifies the pages. Scrolling
in rolling up down the document. It is used when the
user requires scanning individual’s lines.
Messages and comments:
Messages and comments are the systems way of communicating with the
user. They are used for the following purposes.
• Indicate the status of proceeding.
• Indicate the error has been detected.
• Request the user to select an action.
• Verify that the selected action is correct.
Verification messages are used when
• Large amounts of data will be deleted
• Master file records are deleted
• Long running process is initiated
• When a termination of a process or exit from the system is required
• When a termination of a connection with another location is required.
Navigation:
Users may become lost or deeply nested menus may hinder navigation,
hence special are should be taken in navigation issues.
Guidelines:
• Use windows when information from different sources has to be viewed
simultaneously.
• Use logical sequences in menus.
• If only one part of display changes then display only the part.
Key Assignment:
Functions key assignment must be identified. The following functions
can be assigned keys:
• Invoke an action
• Undo an action
• Data storage
• Escape
• Macro command
• Help
• Cursor movement
• File positioning
Guidelines for key assignment:
Assign only one key for a selection
• The function invocation must not require pressing the enter key.
• Function key usuage must be consistent.
• Use common sense when assigning the keys.
• Help Systems
• Messages and comments inform the users whereas the help system instructs
the users. The help system can be designed in two ways:
• Use an index of terms and keywords.
• Use a dialogue to instruct the user in step-by-step procedure.
Guidelines:
Assign a common key for help.
Provide help for the entire system
Provide help only on how to use the system and how to do the job.
2.4File and database development:

Basic file terminology:

Data item: Invidual elements of data are called as data items. For example
the bank cheque consists of data items check number, account number,
amount, date, payee etc. each data item has a name and a specific value
associate with it. Data items can comprise sub items or sub has a name and
a specific value associated with it. Data items can comprise of sub items or
sub fields.

Example address consists of door-no, street, city, pin or date consists of day,
month, year.
Record:
The complete set of related data pertaining to an entry, such as a
customer is known as record. This record is treated as a single unit.
Each field in a record has a defined length and type. When the
number and size of data items in a record are constant for every
record, the record is called as a fixed length record. In the case of
variable length record the record size varies. This may be due to
individual data items being of variable length or a variable no data
items.
Record key:
One data item from a record is selected to distinguish one
record from another. This is called as the record key or key. The key
is already part of the record .Examples are the social security
number in a customer record, or an employee number in an
employee details record.
Entity:
An entity is any person, place or thing, or event of interest about which data
are captured, student, course, department entities in a college information
system.
File:
A file is a collection of related records. Each record in a file is included
because is pertains to a same entity. For example a file of check consists of
many check records. The number of records in a file determine the size of the
file.
Database:
A database in an integrated collection of data stored in different types
of records, so that it is accessible to multiple applications. For example in a
university database records of students, faculty and courses are stored and
these different types of records are interrelated. The relationships here are the
logical relational ship between data.
Data structure diagrams:
Data structure diagrams are graphic tools that show the logical data structure requirements of an
information systems application. They serve the following purposes:
• Verify information requirements
• Describe data associated with entities
• Show the relationship between entities.
• Communicate the data requirements to file manager
• The entities are represented as rectangles, with the entity name on top and a list of attributes
describing the entity. The key is the first attribute listed.
Types of files:
Master file:
A master file is a collection of records about an important aspect of an
organization’s activities.
It may contain data describing the current status of specific events or many contain a history of
events affecting a particular entity .The master files must be kept up to date and accurate.
Transaction file:
Transaction file is temporary file with two purposes.
Accumulating data about events as they occur.
Updating the master files to reflects the results of current transactions.
Master files are permanent and exists as long as the system exists, but their contents
changes as a result of processing and updating. Transaction files on the other hand are
temporary. They are destroyed once when they are no longer necessary.
Table file: it is a special type of master file. They contain reference data
used in processing transactions, updating master files or producing
output. These files store reference data that otherwise would be stored
in master files or would be embedded in programs
Report file: the CPU generally produces data for output at a faster rate
than the printer can possibly keep with. Report files are temporary files
used when printing time is not available for all reports produced. The
data is stored in a report file until it can be printed .This process is
known as spooling.
Other files: some other kinds of files are the backup files and archive
files. Backup files are copies of master or transaction files to ensure
that a duplicate is available if any thing happens to the original file.
Archive files are used for long time storage of data and are generally
used to store history information.
2.5 File Organization:
Sequential organization:
In a sequential file records are stored one after the other
without concern of the value of the data in the records. The first record
written into the file is stored in the beginning of the file and the next
record written is stored in the second position and so on. A sequential
file is read from the beginning. If we want to access the 10th record we
cannot go directly to the lOth record. The record can only be accessed
by starting from the beginning.
Similarly if we need to search for a specific record using some data field
as a key then, the only way possible is by reading the records one by
one from beginning of the file and checking the record for the key
value. When the requirement is to access one particular record in large
files then sequential files are disadvantageous, but we want to access
nearly every record in the file then sequential flies are advantageous. In
general if fewer than 10% of the records are needed the file should not
be organized as a sequential file. But if more than 40% of the records
will be needed then sequential files can be used.
Direct access organization.
Direct access files are keyed files; they associate a record with
a specific key value and a particular storage location. These files do not
require the system to start from the first record. Direct addressing in
this method the record key is itself used as the storage address. For
example if the record key is 1002. Then the record key is stored in the
memory location 1002. This method is simple and quick, but the
requirements for this method often prevent its use.
The requirements for direct addressing:
• The key should be in a dense ascending order with few unused
values. Few open gaps in key values are wanted.
• The record keys correspond to the numbers of the storage
addresses is there is a storage address for each actual or possible
key value in the file and there are no duplicate key values.
Another problem in the use of direct addressing is when the keys of the
records do not match storage addresses.
Hash addressing: when direct addressing is not possible but direct
access is necessary then hashing can be used. Hashing also known as
key transformation is the process of deriving a storage address form the
record key. An algorithm is developed to change the key value into
another value which is used as the storage address.
Types of hashing algorithms:
• Division by prime number: here the key is divided by a prime number and the remainder is used as the
address.
• Folding: here the key is split into pieces and is processed further.
• Extraction: here specific digits are selected from the key and are processed with the remaining digits.
• Squaring: here the key is multiplied with itself and the result is processed further.
• Requirements for hashing algorithms:
•
• Repeatable
• Even distribution
• Minimizes synonyms (synonyms occur when two different keys result in the same storage location after
being processed by the hashing algorithm)
• When a record is to be stored the hashing algorithm is performed and the
record is stored in the resulting location. If the location is already occupied by
another record then the record is stored in separate overflow area. When a
record has to be retrieved the key value is got and the hashing algorithm is
performed. The record is got from the storage location got from the algorithm. If
the record is not found in the location then the overflow area is searched.
When a record is to be stored the hashing algorithm is performed and the
record is stored in the resulting location. If the location is already occupied
by another record then the record is stored in separate overflow area.
When a record has to be retrieved the key value is got and the hashing
algorithm is performed. The record is got from the storage location got from
the algorithm. If the record is not found in the location then the overflow
area is searched.
Indexed organization:
The index is a separate file from the master file. It includes a record key and the storage address for that
record. To find a specific record the index is first searched to find the key of the record wanted. When it is
found the corresponding storage address is noted and the record is accessed directly. There are two types
of indexed organization.
Indexed non-sequential organization
Here the master file is not in any specific order. There is one entry in the index for each record in the file.
The records in the master files are stored in a random order.
Index

Rec-key Block addr

AA 1021

AB 1018

CC 1024
 Indexed sequential organization

AB ……. AA …….. CC

1018 1021
1024

The groups of records are stored in blocks. The records inside each block are in sequential
order. Not all records in the file are adjacent to each other and in sequential order.

The adjacent blocks are also not in sequential order. Only one entry per block in needed in
the index ie the highest record key in the block is stored in the block appears in the index.

When a record is to be retrieved. When is found the address of the block is used to access
the block and the needed record is searched for inside the block.
 Index
Rec-Key Block addr
1115 1345
1315 1349
1429 1346

1345

101 101 101 111

0 1 5 5

1346
131 …. 142
6 8

1349
111 …. 172
7 5
Systems development in a database environment

Relationships among data

In transaction processing the focus is on only one entity, but as more new applications emerge there is
need to focus on the relationships between different entities.

For example an order processing system may need three entities order, customer and item. Here a
single instance of an order entity is related to the customer entity and the item entity (the customer who
placed the order and the items ordered).

Thus database management is about:

● Drawing on the natural relationships in the data.

● Sharing data about the entities across all applications that need data.
These relationships among data are drawn from the real world relationships of the entities they
represent. These entities and their relationships can be represented graphically with the help of an entity
relationship diagram.

Entity relationships are described by their dependence on each other and on the extent of the
relationship.
Entity dependency

There are two common types of dependencies:

Existence dependency: one entity is unable to exist in the database unless the other is first
present. For example orders do not exist unless there is a customer. Eliminating records from one
entity may result in elimination of records of other entities if there is an existence dependency.

Identification dependency: this is the case where an entity cannot be uniquely identified by its own
attributes, but only through its relationship with other entities.

Extent of dependency

The extent of the dependency is based on:

● The direction of the relationship

● The type of association between them
Data structure diagrams

The basic components of a data structure diagram are:

● Entities
● Attributes
● Attribute pointers
● Logical pointers
Attribute pointers link two entities by common information, usually a key
attribute in one and a non key attribute in another.

Logical pointers identify the relationships between entities. In the

following diagram there are three entities; order, customer and item. The
customer no and the item number in the order entity are the attribute
pointers. The arrows connecting the entities represent the logical
pointers. The crow’s foot in the arrow shows a one too much relationship,
for example one order can have any number of items and one customer
can have any number of orders.
ORDER
Order no
ITEM
Item desc
Item no
Qty
Item desc

price
Cost
Customer no

Item no
CUSTOMER

Customer no

Cust name

Address
Sharing data across applications

Different systems may need to access the same data. So the database
should be designed so as to share the data between all the
applications that need to access the data. The main reasons for
sharing the data are:

Redundancy: if data is not shared then the same data has to be

stored by the different applications. This only leads to waste of
storage space.

Integrity: if same data is stored in different master files, it may

lead to loss of integrity of the data i.e. The data in the different files
may not agree. This is because some updations may have been made
on some of the master files and not on others.
The impact of DBMS on system design:

The use of database management does not eliminate the need for computer
programs. The DBMS is a bridge between the application program that determines what
data are needed and how they are processed and the OS, which is placing the data in
storage. The schema is the database and the subschema is the portion of the database
that the application will use. These are the steps taken to retrieve data from the
database:

⮚ The application program determines what data are needed and communicates the
need to the DBMS.
⮚ The DBMS determines if the data is present in the database.
⮚ The Dbms instructs the OS to locate and retrieve the data.
⮚ A copy of the data is given to the application program.
Data structures for interrelating data:

Multilist: A multilist links common items together in a

file. It acts like a chain, with each link a record that meets some
specific criteria. The records are not physically moved around,
but all records that have a specific attribute are logically linked
together. we can have many lists, for example in an order
processing system we can have lists based on order no, item
code or customer etc.,
 Inverted file: here an index is used to store information about the
location of records having particular attributes. In a fully inverted file there is
one index for each type of data item. The invertered file data structure for
order.

Each record in the index contains the storage address of each record in
the file that, meets the attribute. Some attributes will not be used to retrieve
data and hence they are not indexed then the file is a partially inverted file.
Some databases may also include count information, ie the information on how
many records meet the particular requirement.
 Datamodels:

There are three widely used

Item index location
data models:

▪ The relational model

23 1344,1455
▪ The hierarchical model
24 6755,7655,7899
▪ The network model
45 8976,8766
56 3454
78 4563
2.6System design in a database environment:

In most organizations a separate database management staff

oversees the design and the development of the databases. The
database management staff defines the schema, maintain the
datadictonary and enforce standards for the data. The system analyst
must define the requirements and program processing specifications,
which are translated to the sub schema. The steps are involved
when there is an existing database is:
⮚ Become familiar with the database schema
⮚ Review the data standards and specifications in the data dictionary.
⮚ Determine the logical data requirements and develop the
subschema
⮚ Identify and design any of the master or transaction files that are
needed but not in the database.
⮚ Determine the proper identification, coding, validation schemes and
processing procedures for all entities in the database and fields.
⮚ When a new database has to be developed then all the above steps
are necessary and in addition there must be interaction with the
database development staff for the establishment of the schema
,data definitions and the data dictionary.
2.7Design in a distributed environment:

A distributed system that connects locations that have computer capability to

capture and store data to process data, and information to other systems. The range
of computing capability varies. There is no requirement that all the equipment must
from the same manufacturer. There are several make a autonomous hardware
involved.

All nodes have ability to capture and process data where events are occurring.
They can also transmit data or reports from their system to another system through a
communication network.

For example large organizations establish networks to interconnect multiple sales

and manufacturing locations .thus sales, accounting and technical information are
readily accessible for multiple locations. The computer management, as the top of the
network will be able interacts with other locations while having its own information
Characteristics of distributed systems:

▪ Communication is the backbone of distributed systems: the other

features are:
▪ Multiple general purpose processing components: the processing systems
can be assigned specific jobs on a dynamic basis.
▪ High-level operating system: individual processing systems have their
own operating systems, but there is also higher –level network operating
system.
▪ Physical distribution of components.
▪ System transparency: users do not know anything about the location,
make, speed, OS etc of the components. They are referred by their
names.
▪ Dual components roles: individual processing components can operate
independently or can act as an integral part of the network.
The following are excluded from the classification as distributed
systems.

⮚ A multifunction mainframe that distributes processing

between separate input/output and peripheral processors.
⮚ A front-end processor that controls communications
functions for the system to which it is attached.
⮚ A collection of remote terminals that collect and transmit
data to the host system.
⮚ The interconnection of multiple host computers that transmit
messages and perform dedicated functions tasks.
⮚ A personal computer that can be portioned.
Reasons for designing distributed systems:

Local processing with communications capability: Distributed systems are

designed when it is necessary to capture and process data locally but also to
communicate with other locations. For example data from transactions during the
day can be captured and stored locally and the summary reports of the
transaction can be sent at the end of the day to the head quarters.

Linking different makes of equipment: the brand of the systems or the

systems can be vary between the nodes. Even the software can be different .In
this way the local operations can be tailored to their own unique characteristics.

Sharing loads: Work can be shared between the sites .An added advantage is
reliability, ie even when one node is down the work can be done by the other
nodes in the system.
Sharing software: Some software will run only on certain
makes of equipment or budget constraints may prohibit all
the sites from purchasing some expensive software.
Software sharing permits a remote users to access the
computer system at another node, enter data and have it
run on the remote using the s/w stored on that system.

Example:

Refer the banking (credit card) example for distributed

system.
Summary:

▪ Entity:
An entity is any person, place or thing, or event of interest about which data are
captured, student, course, department entities in a college information system.

▪ File:
A file is a collection of related records. Each record in a file is included because is
pertains to a same entity. For example a file of check consists of many check
records. The number of records in a file determines the size of the file.

▪ Database:
A database in an integrated collection of data stored in a different types of records,
so that it is accessible to multiple applications. For example in a university database
records of students, faculty and courses are stored and these different types of
records are interrelated. The relationship here is the logical relational ship between
data.
There are three widely used data models:

● The relational model

● The hierarchical model
● The network model
RELATIONAL MODEL
Relational Data Model
Relational data model was developed by E.F. Codd in 1970. There are no physical links as they are in the
hierarchical data model. Following are properties of the relational data model :

● Data is represented in form of a table only.

● It deals only with data, not with physical structure.
● It provides information regarding metadata.
● At the intersection of row and column there will be only one value for the tuple.
● It provides a way to handle queries with ease.