Systems Analysis

  1. What is process modeling?
  2. How to use data flow diagram (DFD) and data dictionary for process modeling?
  3. What is logic modeling?
  4. How to use structured English, decision tables, and decision trees for logic modeling?
  5. What is a systems proposal and its role in systems analysis?

 

Analysis II -- Requirements structuring

To organize information gathered during requirements determination into a meaningful representation of process, data, and logic views of the information systems

 

Process modeling

To represent the functions/processes which capture, manipulate, store, and distribute data between a system and its environment and between components within a system, i.e., what are involved in converting data into information?

 

Data flow diagram

A technique for graphically depicting, at levels of increasing detail, the transformation of data into information by processes

Gane & Darson’s 4 symbols (Figure 9.1)

1.  Data flow -- data moves from one place of the system to another

2.  Data store -- data at rest

3.  Process -- work/actions performed on data

4.  External entity -- origin/destination of the data

 

Steps in developing DFDs (Figure 9.2)

  1. List business activities to identify processes, external entities, data flows, and data stores
  2. Create a context diagram
  3. Create the next level diagram
  4. Create child diagrams
  5. Check for errors
  6. Develop a physical DFD

 

An example: UK bookstore cap and gown order processing

Step 1:  A list of business activities

1.      Students place orders by filling out a form.

2.      Order department receives orders by mail, fax, or personal delivery.

3.      Order department processes orders by verifying that all order information is accurate and that the item ordered is currently available in stock.

4.      Information from valid orders is used to update the student and item master records.

5.      Valid orders are forwarded to the shipping department to be filled.

6.      A receipt is produced notifying the student the status of his/her order.

 

Step 2: Context diagram

 

 

Step 3: Level-0 diagram

 

 

Step 4: Level-1 diagram


Step 5: Check for error (Figures 9.5 and 9.6)

DFD rules

Internal consistency rules

Elements

Rules

DFD

·        At least one process

·        No more than 9 processes

Context diagram

·        Contains only one process numbered 0

·        At least one input from an external entity and one output to an external entity

External entity

·        Appears only on the context diagram

·        Connected to a process

·        Labeled with noun phrase

Process

·        At least one input data flow and one output data flow

·        Inputs to process are different from outputs of that process

·        Labeled with verb phrase

Data flow

·        Has only one flow direction

·        No split

·        No loop

·        Labeled with noun phrase

Data store

·        An interface between two processes

·        Labeled with noun phrase

 

Hierarchical consistency rules

Elements

Rules

DFD

·        A parent diagram must exist unless it is a context diagram

Process

·        Decompose to either another diagram or a primitive process specification

·        Numbered with respect to its parent

Data flow

·        An input (output) data flow on a parent diagram must appear on a child diagram as input (output)

·        An input (output) data flow on a child diagram must appear on a parent diagram as input (output)

Data store

·        Decompose to either a file definition or a record definition

 

 

 

Step 6:  Develop a physical DFD

Logical vs Physical DFDs (Figure 9.8)

Features

Logical

Physical

Model

How the business operates

How the system will be implemented

Process

Business activities

Programs, program modules, manual procedures

Data store

Collections of data

Physical files and databases, manual files

Type of data store

Permanent data collections

Master files, transaction files

System controls

Business controls

Controls for data validation, record status, system security

 


Data dictionary

A documentation and reference of the metadata: data on

  1. Data flow (Figure 10.3)
  2. Data structure (Figure 10.5)
  3. Data elements (Figure 10.7)
  4. Data stores (Figure 10.12)

 

à to promote understanding of the data of the system by collecting, coordinating, and confirming what a specific data term means to different people in the organization.

 

Data types

Specific description

Data flow
  • Source
  • Destination
  • Type (File, Screen, Report, Form, Internal)
  • Data structure
  • Volume

Data structure
  • =
  • +
  • { }
  • [  ]
  • (  )
  • is composed of
  • and
  • repetitive elements
  • either/or
  • optional

Data element
  • Alias
  • Length
  • Type (alphabetic, alphanumeric, date, numeric)
  • Input/Output format
  • Default
  • Base/Derived
  • Continuous/Discrete
  • Validation criteria

Data store
  • File type
  • File format
  • Record size
  • Number of records
  • Growth rate
  • Primary key
  • Secondary keys

 

 

 

Data flow description example:


Logic modeling

To represent the internal structure and functionality of the processes represented on DFDs, i.e., how do data be converted to information?

 

Structured English

relies on action verbs and noun phrases without adjectives or adverbs to specify 3 typical logic in structured programming: sequence, selection, and repetition

sequence: sequential order of the statements

selection: IF_THEN_ELSE; SELECT_CASE

repetition: DO_UNTIL; DO_WHILE

 

Cap & gown ordering systems example

Process 3.1  Validate Student Status

MATCH Student_Information with Student_Record using student’s Last_Name and First_Name

BEGIN IF

     IF Student_Not_Found

     THEN RETURN Student_Not_Found

ELSE

            IF graduation_date is not equal to May, 2001

            THEN  RETURN Student_Invalid_Status

END_IF

RETURN Valid_Student

 

Process 3.2  Validate Order Item

MATCH Order_Item_Information with Available_Item_Record based on Item_Description     

BEGIN IF

IF the Order_Item_Quantity is grater than the Available_Item_Quantity

            THEN RETURN Insufficient_Quantity

END_IF

RETURN Valid_Order

 

Decision tables

all possible choices and conditions the choices depend on are represented in tabular form: condition, actions, and rules

Conditions and Actions

Rules

 

Conditions

 

 

Condition alternatives

 

Actions

 

 

Action Entries

 

Maximum number of rules in the table =

n: number of conditions

Ci: number of alternatives for condition i

 

 

Conditions and Actions

Rules

1

2

3

4

Under $50

Y

Y

N

N

Pays by check with 2 forms of ID

Y

N

Y

N

Uses credit card

N

Y

N

Y

Ring up sales

X

 

 

 

Decline sales

 

X

 

 

Call supervisor for approval

 

 

X

 

Call bank for credit authorization

 

 

 

X

 

Cap & gown ordering system example:

Process 3. Validate Order

Conditions/Actions

Rules

Student found

Y

-

N

-

Graduate in May

Y

N

-

-

Item in stock

Y

-

-

N

Accept order

X

 

 

 

Reject order

 

X

X

X

 

Decision trees (Figure 11.22)


a decision or choice situation is depicted as a connected series of nodes (decision points) and branches (decision alternatives)

 

 

Cap & gown ordering system example

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


When to use what?

Condition

Recommendation

Many repetitious actions

Structured English

Communication to end users is important

Structured English

Complex combinations of conditions, actions and rules

Decision tables

Checking for redundancies, contradictions, possibilities

Decision tables

The sequence of conditions and actions is critical

Decision tree

Not every condition is relevant to every action

Decision tree

 

Evolution of systems analysis techniques

Era

Orientation

Techniques

Evaluation

Pre-computer

Flow-oriented, i.e., no logical details

·        Process flow charts

·        Forms flow charts

·        Little clarification of system structure

·        Weak on the treatment of procedure

1st Generation (1950's)

Flow-oriented

·        System flow charts

·        Flowcharts

·        Message specification sheets

·        Reduce readability to non-technical people

·        Human is considered external to the system

2nd & 3rd Generation (1970's)

Package-approach

·        Decision tables

·        Gridcharts

·        Analysis packages, e.g., IBM's SOP (Study Organization Plan)

·        Multiple system representations require multiple tools

4th Generation

(1980's)

Structured

·        Functional decomposition

·        IBM's HIPO

·        DFD

·        Data dictionary

·        Structured process specification

·        Improved consideration of system structure issues

 

 

Findings

·        A well structured implementation of a poorly designed system yield few benefits

·        A well designed solution to a poorly understood problem provided few rewards

·        No single tool, technique, or methodology can support the complete analysis of today's complex systems

·        DFD is easier to learn and easier to use than functional decomposition.