Practical Significance of DFDs
 Many older information systems (legacy systems):
 Are mainframe-based (inaccessible to most non-IT users);
 Use non-relational (hierarchical) databases, and
 Require second-generation programming languages (e.g.
COBOL) to perform data input, queries, reporting.
Thus, many IT professionals focus on data flows and
physical implementation, not data relationships or RDBM
design principles.
 Many business processes are so complex that E-R design
alone will not give analysts insight into their businesses
DFDs allow the analyst to determine: What does the
organization do? How does the organization do it?
1
Creating DFDs
Define Entities
• External entities represent
persons, processes or machines
which produce data to be used
by the system or receive data
that is output by the system
• Examples: Student, Customer,
Client
Define Processes
• Processes are discrete actions
that transform input data to
output data
• Examples: Create Student
Record, Calculate Purchase
Cost, Register Client
Student
2.1
Create
Student
Record
2
Creating DFDs (cont’d)
Define Data Stores
• Data stores are temporary or
permanent repositories of
information that are inputs to or
outputs of processes
• Examples: Student Master, Client
List
D3
Student
Master
Define Data Flows
• Data flows represent the transfer of
data over time from one “place”
(entity, process, data store) to
another
• Examples: New Student Information
(from Student, to Student Master)
New Student
Information
3
Creating DFDs (cont’d)
Define the System
• A system is the collection of all
business processes which perform
tasks or produce outputs we care
about. It is “what happens.”
• The system is a single process,
connected to external entities
• Represented in a “Context
Diagram”
(Figure 4.13,
Shelly,
Cashman and
Rosenblatt)
Define Subsystems
• A subsystem gives a more
detailed view individual processes
contained in the context diagram
• Includes data stores, more
elementary processes
4
DFDs Created by Top-Down Analysis
 Create a narrative: description of
Narrative
system
 Create a Context Diagram that
contains a single process (“the
system”) and all entities which share
data with the system
Context Diagram
Diagram 0 DFD
 Explode the “parent” context diagram
to produce a Diagram 0 (“child”) DFD
 Create Diagram 1, 2, …, n DFDs that
Diagram 1 DFDs
represent “explosions” of Diagram 0,
1, …, n-1 DFDs until a diagram has only
“primitive” processes
 Create process descriptions to be
implemented by application
programs: queries, macros, reports,
programming languages
Diagram n DFDs
E-R Diagram
Process Descriptions
5
Where to Begin Creating DFDs
 Start with the data flow from an external entity and work
forwards
 Start with the data flow to an external entity and work
backwards
 Examine the data flows into or out of a data store
 Examine data flows, entity connections and data stores
associated with a particular process
 Note fuzzy, ill-defined areas of the system for further
clarification
6
What to Avoid in DFDs
4
Processes with no outputs or
no inputs
Perform
Repair
Processes whose inputs are
obviously inadequate to yield
outputs
Connecting data stores directly
to each other
Class List
Students
Courses
Having data flows terminate at
data stores
Connecting entities to anything
other than processes
Payroll
Department
Employees
Making the data flow diagram too
cluttered (e.g.  9 processes)
Many processes with a single
input and output (linear flow)
1
2
3
Process
A
Process
B
Process
C
7
DFD Example: Bus Garage Repairs
Buses come to a garage for repairs.
A mechanic and helper perform the repair, record the reason for the repair and
record the total cost of all parts used on a Shop Repair Order.
Information on labor, parts and repair outcome is used for billing by the
Accounting Department, parts monitoring by the inventory management
computer system and a performance review by the supervisor.

Key process (“the system”): performing repairs and storing information
related to repairs

External Entities: Bus, Mechanic, Helper, Supervisor, Inventory Management
System, Accounting Department, etc.

Processes:





Record Bus ID and reason for repair
Determine parts needed
Perform repair
Calculate parts extended and total cost
Record labor hours, cost
8
DFD Example: Bus Garage Repairs
(cont’d)
 Data stores:




Personnel file
Repairs file
Bus master list
Parts list
 Data flows:
 Repair order




Bus record
Parts record
Employee timecard
Invoices
9
Bus Garage Context Diagram
Bus
Fixed
mechanical
problems
Mechanical
problem
to be repaired
Helper
Labor
0
Bus
Repair
Process
Labor
Mechanic
Labor,
parts cost
details
Repair
summary
List of
parts used
Supervisor
Inventory
Management
System
Accounting
10
Bus Garage Diagram 0 DFD
11
Bus Garage Diagram 0 DFD
(cont’d)
12
BIS 360 – LECTURE SEVEN
Process Modeling
(Chapter 8)
Today’s Lecture Outline








Where are we?
Why Process Modeling?
What is Process Modeling?
Basic Symbols for Process Modeling
System Concept vs. Process Decomposition
Decomposition Rules
Basic Concepts about a Data Flow
Rules for Data Flow
Where are we?
Project ID and Selection
1. Determine system requirements
2. Structure system
requirements (ch. 8-10)
Project Initiation
& Planning
3. Generate alternatives for selection
Analysis
Logical Design
Physical Design
Implementation
Data Modeling
Process Modeling
Logic Modeling
Why Process Modeling?
An Information Engineering Approach
Generic IE Approach:
Data
Model
Info
System
Process
Model
Process
enrollment
STUDENT
Students
SECTION
Registration
ADVISING
FACULTY
COURSE
Assign
Advisor
Process
registration
Courses
Student
roster
Section
Advisees
Listing
Transcript
What is Process Modeling?
Process modeling - a technique to organize
and document the system structure and data
flow between system’s PROCESSES and their
relevant procedures to be implemented by a
system.
Data Flow Diagramming  Another logical modeling tool to support the
process modeling
 the second step in the Analysis stage of SDLC
Basic Symbols of Process
Modeling
Data Flow Diagram and Its Components
DFD: A diagram about the data flow between
external agents (sources/ sinks) and the
processes and data stores within a system
Key Components:
External
Agent
An Example A Data Flow Diagram for a Banking
System
An Information System
A Generic View
In general, a system could be viewed as a single
Process
Source
(Customer)
Trans
Data
0
Information
System
Report
Sink
(Mgmt)
You could have multiple sources and sinks!
This generic diagram is called “Context Diagram”
A Context Diagram
 An overview of an organizational system that
shows the system boundary, sources / sinks
that interact with the system, and the major
information flows between the entities and
the system
 A Context Diagram addresses only one
process.
 An example ...
An Example A Context Diagram for a Fast-Food IS
A Systematic Way for Process
Modeling
Process Decomposition
In general, a system could be too complex to
understand when viewed as a single Process
 We need a Process Decomposition scheme
i.e., to separate a system into its subsystems
(sub-processes), which in turn could be further
divided into smaller subsystems until the final
subsystems become manageable units (i.e.,
primitive processes!)
A divide and conquer strategy!!
Level-0 Diagram
 A DFD that represents the primary
functional processes in the system at
the highest possible level
 An Example ...
An Example - A decomposed Context
Diagram - Level 0 Diagram
An Example - A further
decomposition
A Level-1 Data Flow Diagram
Process Decomposition Rules
Generic Decomposition Rules:
 A process in a DFD could be either a parent
process or a child process, or both.
 A parent process must have two or more child
processes.
 A child process may further be decomposed into a
set of child processes.
Decomposition Overview
Context Diagram
Level-0
Diagram
Three Major Types of Process
 Function Process - A function is a set of related
activities of the business (e.g., Marketing,
Production, etc.)
 Event Process - An event process is a logical unit
of work that must be completed as a whole. (e.g.,
Process customer credit verification)
 Primitive Process - a primitive process is a
discrete, lowest-level activity/task required to
complete an event. (e.g., Check the credit card
balance)
Naming Rules for Processes
 Function Process - use a Noun!
 Event Process - Use a general action verb
 Process Student registration.
 Respond to ...
 Generate ...
 Primitive Process - use a strong action verb
 Validate Student ID
 Check ...
 Calculate ...
Rules for Processes
 No process can have only outputs (a miracle!)
 No process can have only inputs (a black
hole!)
 No process can produce outputs with
insufficient inputs ( a gray hole!)
Can You Identify Errors in This
Diagram?
What’s wrong?
Membership
application
Employee
Bank statement
Black Hole
Existing account
3.1.1
Generate an
Employee bank
statement
3.1.2
Create a new
member account
Employee
status
Employee address
Member Accounts
New account status
Employees
Miracle
3.1.3
Freeze member
account number
Frozen account notification
Accounts
Receivable
Department
Gray Hole
Processes in a DFD
Correct vs. Incorrect
Incorrect
Correct
Basic Concept About Data Flows ...
It has two kinds of flow: a) Inflow to a Data Store
(Create/Modify/Delete) b) Outflow from a Data Store (Read)
Order
Cancelled Order
Delete
Process
Order
2
2
New
Order
Cencel
Order
Order
to be
Deleted
Read
Orders
Unfilled
Order
l
Create
New
Order
Address
Change
Order
Address
Change of Address
Modify
1
2
Summarize
Unfilled
Orders
Summary of Orders
Rules for Data Stores
 Data cannot move directly from one data store to
another data store -- it must be moved by a process.
 Data cannot move directly from an outside source to
a data store -- it must be moved by a process.
 Data cannot move directly to an outside sink from a
data store -- it must be moved by a process.
 You need to use a Noun phrase to label each data
flow
Data Flows in a DFD
Incorrect vs. Correct
Correct
Incorrect
Naming Rules Data Flow
 Use a singular noun phrase for each data
flow
 Ex: customer data, shipping report, …, etc.
 Carry logical meaning only, i.e., no
implication on data form or data structure
 Minimum flow (no data flooding!!)
 Should never be “Unnamed!!” - otherwise,
there might be a modeling error.
Naming Scheme for Other DFD
Components
 Process (Event) - Use an Action Verb Phrase
 Process member order, Generate bank statement, ...
 External Agent (Sink/Source) - Use a singular
descriptive noun
 Ex: Student, Customer, etc.
 Data Store –
 Use a plural descriptive noun (Members, Customers,
etc.)
 Or use a noun + file (Inventory file, Goods sold file)
Basic Rule in
DFD Decomposition
Balancing Principle -- the decomposed DFD
(I.e., the next lower level DFD ) should retain
the same number of inputs and outputs from
its previous higher level DFD (I.e., No new
inputs or outputs when a DFD is
decomposed)
Balancing Principle
Context Diagram
Level-0
Diagram
Unbalanced DFD