An Introduction to

Systems Analysis
and Modelling
Dr Kristin Stock

Outline
What, Why, When and Where?
Different levels of modelling (conceptual, low level

etc.).
Modelling and Design what is design, how does it
relate to what we will cover in this course?
The systems development lifecycle.
Systems development methodologies.
Systems analyst roles and skills.
Project management, especially re CASE Tools.

What is Systems Analysis

and Modelling?
How do

programmers
know what to write
in their code?

What is Systems Analysis

and Modelling?
Systems analysis and modelling is a process

of:
identifying the problem that needs to be

solved
analysing the current situation
identifying requirements for the new system
modelling those requirements in a systematic
way
creating a clear specification of what the
system must do

What is Systems Analysis

and Modelling?
When we talk about specifying what the

system must do, specifically this can include:

What data is needed
The rules that govern that data
The rules for how that data will be processed
The layout of user interfaces, reports, etc.
Rules about access constraints
Expectations about the performance of the

system

Is it about designing a
system?
Well..yes, but

there are
different
definitions of
design.

What is the difference between

analysis, modelling and design?
Analysis is problem focussed
Design is solution focussed
But one morphs into the other through the

process.
The tools we use for analysis are also used for
design.
We model the existing process, and then we
model our design of the new process.
We can sometimes use this to generate code

Why Systems Analysis and Design?

To ensure the system meets the

needs it is intended to meet.

There is a long history of
expensive IT systems that failed to
meet user needs.
To ensure system quality.
To ensure a good fit with the IT
environment
Ultimately to provide value to
the organisation

Examples of Project Failures

National Health Service, UK
http://calleam.com/WTPF/?p=2003
Transit Ticketing Authority, Melbourne,

Australia
http://calleam.com/WTPF/?p=455

Department of Environment, Food and Rural

Affairs, UK
http://calleam.com/WTPF/?p=1881

The
The systems
systems
analyst:
analyst:
translates,
translates,
connects
connects
communicate
communicate
ss

bridges
bridges the
the gap
gap
between
between the
the
users/organisati
users/organisati
on
on and
and the
the
technological
technological
details
details

When?
Right at the beginning!
When the project idea

is first formed, some

systems analysis
activities are
appropriate.
Can continue
throughout, depending
on SDLC methodology.

Where?
The systems analyst needs to spend

time with the users.

Need to observe and understand the
current situation, systems and
environment.
This is the foundation for a good
system specification.

Different Levels of
Modelling
Conceptual
Increasing levels of detail
Logical

Physical

Conceptual Analysis and Modelling

High level.
Sketches out broad concepts.
Can work with users to define high level

concepts.
e.g. Concept Mapping.
is in
house

is in
street

city

Logical Analysis and

Modelling
Much more detailed.
Independent of implementation.
Does not depend on a particular operating

system, software or hardware.

Implementation decisions can be made
separately, later.
Changes in the implementation do not affect
the logical design.
PIM = platform independent model

Physical Analysis and Modelling

Moves from logical design into design on a

particular operating system, hardware,

software.
E.g. Creating the actual database tables,
creating software classes etc.
Some tools are available to generate some
parts of the physical design from the logical
design models.
E.g. logical database design can be used to
generate specific Oracle, SQL Server
databases.

Systems Development
Life Cycle (SDLC)

The SDLC Process

The process consists of four phases
Each phase consists of a series of steps
Each phase is documented

(deliverables)
Phases are executed sequentially,
incrementally, iteratively or in some
other pattern

Questions to be
Answered
Planning phase

Why should we build this system?

What value does it provide?
How long will it take to build?

Analysis phase
Who will use it?
What should the system do for us?
Where & when will it be used?

Design phase
How should we build it?

Increasing
levels of
detail

SDLC: The Planning

Phase
1. Project Initiation
Develop/receive a system

request
Conduct a feasibility analysis

2. Project Management
Develop the work plan
Staff the project
Monitor & control the project

Deliverabl
e: Work
Plan

SDLC: The Analysis Phase

1. Develop an analysis strategy

Model the current system

Formulate the new system

2. Gather the requirements

Develop a system concept

Create a business model to

represent:

Business data

Business processes

3. Develop a system proposal

Deliverabl
e: Analysis
and HighLevel
Design

SDLC: The Design Phase

1. Develop a design strategy
2. Design architecture and interfaces
3. Select hardware, software, network

infrastructure)
4. Develop databases and file

specifications
5. Develop the program design to

specify:

What programs to write

What each program will do

Deliverabl
e: Detailed
Design
Specificati
on

A word about
specifications
Requirements specification.
What is required of the system?
High level.
Functional specification.
What functions must it provide?
What business processes must it support?

Design specification.
Details of design, including user interfaces,
what each button does etc.
Do we need all three?

SDLC: The Implementation

Phase
1. Construct the system
Build it (write the programming code)
Test it

2. Install system
Train the users

3. Support the system

(maintenance)

SDLC: Methodologies
Methodology: a formalized approach to

implementing the SDLC

Categories
By area of focus
Process

oriented
Data centered
Object-oriented
By sequencing and emphasis of phases:
Structured
Rapid action development
Agile development

Classes of
Methodologies
Structured Development
Waterfall Development
Parallel Development

Rapid Application
Development
Phased
Prototyping

Agile Development
eXtreme Programming
SCRUM

Structured Methodologies
Used commonly in 1980s
Replaced ad-hoc approach.
Formal, sequential, step-by-step.

Waterfall Development

Proceeds sequentially from one step to the next.

Lots of documentation, approved at each step.
Cant go backwards.
Advantages:

Requirements are identified long before programming

starts.
Requirements do not change..

Disadvantages:
If requirements are missed,

expensive postimplementation changes

required.
Does not cater for changes in
the environment.

Parallel Development
After a general high level design, sub-projects

are created.
Sub-projects designed and implemented in
parallel.
Sub-projects integrated at the end.
Advantages:
Reduces time for total development, so less

change in environment.

Disadvantages:
Sub-projects may be interdependent, so

integration is difficult.

Rapid Application
Development
Started to be used in 1990s
Try to get some parts of the system

developed quickly and out to users for

feedback.
Advocate the use of Computer Aided Software
Engineering (CASE) tools, Joint Application
Design (JAD) sessions, visual programming
tools, code generators.
Can result in scope creep requirements
change and increase as users work with the
system.

Phased Development
Begins with a high level design of the system concept.
Defines a series of versions, with ever increasing

functionality.
For each version, there is analysis, design,
implementation, release to users.
Continues until system complete or obsolete.
Advantages:
Gets something in the hands of the users quickly, so early

value to organisation.
Can also identify problems and changed requirements early.

Disadvantages:
Users are working with an incomplete system.
User expectation management.

Prototyping
Iterative.
Very preliminary design used to create first prototype.
Gradually refined.
Analysis, design and implementation performed

concurrently.
All iterative.
Advantages:
Early user involvement and feedback.

Disadvantages:
Such rapid change that methodological design and

analysis is difficult.
Can be very inefficient.

Throwaway Prototyping
Adopts a more

conventional waterfall
process overall, but
prototypes are built along
the way, then discarded.
Prototypes address specific
issues.
Each prototype has its own
analysis, design,
implementation cycle, and
is used to refine the
design.

Agile Development
Most recent.
More focussed around design philosophies

than specific details of methodology.

Follow the 12 principles:
Software is delivered early and continuously
Changing requirements are embraced
Customers and developers work together
Face to face communication is key
Focus on technical excellence and good design
Regular reflection.

Drawbacks of Agile Development

Require co-location of

development team.
The process if not highly
managed (by definition), so can
turn into programmers gone
wild, hacking together
solutions.
Little documentation.
Suitable for large, mission
critical systems?

Extreme Programming
Communication, simplicity, feedback,

courage.
Feedback to users continuously.
Embrace change.
Quality first.
Testing is continuous.
Evolutionary, incremental.
Heavy on-site user input.
Relies on a small, close-knit team.

Scrum
Made up of 30 day sprints.
Each sprint delivers a system to the

user.
No requirements changes during a
sprint.
Scrum meeting each day:
Previous days accomplishments.
Plan for today.
Any obstacles.

Again, may not be suitable for large

projects.

Which Methodology to
Use?

What kind of person is involved

in Systems Analysis?
Agents of change
Identify ways to improve the organization
Motivate & train others

Ability to see both technical and user sides of

a problem.
Big picture plus detail.

Skills needed for Systems

Analysis
Skills needed:
Technical: must understand the technology
Business: must know the business processes
Analytical: must be able to solve problems
Communications: technical & non-technical

audiences
Interpersonal: leadership & management
Ethics: deal fairly and protect confidential
information

Jobs involving Systems Analysis

Business Analyst
Focuses on the business issues
Systems Analyst
Focuses on the IS issues

Infrastructure Analyst
Focuses on the technical issues
Change Management Analyst
Focuses on the people and management issues

Project Manager
Ensures that the project is completed on time and within
budget
Systems Architect
Data Modeller
User Interface/Interaction Designer

IT Project Management
Project Management is the process of planning and

controlling system development within a specified

time at a minimum cost with the right functionality
A project is a set of activities with a specified
beginning and end point meant to create a system
that brings value to the business
Project Managers monitor and control all tasks and
roles that need to be coordinated.
Analysis and Modelling are key to the success of IT
projects.

Spectacular IT project
failures
UK prison IT system:
http://www.zdnet.com/article/uk-prison-it-massiv
e-and-spectacular-failure
/
See reading from Nelson, 2007 on Stream.

Why are failures so

common?
Underestimating time and effort.
Unreasonable demands/expectations.
Use of unknown/new technologies.
Too much focus on technology:
Driven by technology rather than business
value.
Failing to take human, organisational, cultural
aspects into account.

Project Identification
A business need is identified management,

business unit, IT department, external (e.g.

politicians, senior management).
Could result from some problem (losing
market share etc.).
Speculative project identification.
The business need should drive the high level
requirements.

The System Request

A document that describes the reasons for and the

value added from building a new system

Contains 5 elements:
Project sponsor: the primary point of contact for the

project
Business need: the reason prompting the project
Business requirements: what the system will do
Business value: how will the organization benefit from
the project
Special issues: Anything else that should be considered

What is a Business Case?

Commonly (and loosely) used term to

describe a document that justifies why a

system should be built/changed/upgraded.
Can range from a couple of paragraphs to a
lengthy document.
If you want to spend time or money on a
project, you will need to make the case to
management.

Feasibility Analysis
Is this project feasible?
What are the risks?
Can these risks be overcome?
Major components:
Technical feasibility (Can we build it?)
Economic feasibility (Should we build it?)
Organizational feasibility (Will they use it?)

Technical Feasibility
Identify risks in the following areas:
The functional area: Are analysts familiar
with this portion of the business?
The technology: Less familiarity generates
more risk
Project size: Large projects have more risk
Compatibility: Difficult integration increases
the risk

Economic Feasibility
(Cost-Benefit Analysis)
Identify the costs and the benefits
Assign values to the costs and benefits
Determine the cash flow
Determine the value using one or more

methods:
Net present value (NPV)
Return on investment (ROI)
Break-even point

Formulas for Determining

Value

Example Cost-Benefit
Analysis

Example Break-Even
Point

Organizational Feasibility
Will the users accept the system?
Is the project strategically aligned with the

business?
Conduct a stakeholder analysis
Project champion(s)
Organizational management
System users
Others

Project Selection
Projects are approved, declined or delayed

based on value added vs. risks

Project portfolio management
Goals:

Maximize cost/benefit ratio

Maintain an optimal mix of projects based on:
Risk
Size, cost & length of time to complete
Purpose, scope & business value

Limited resources require trade-offs

Selected projects enter the project

management process

Project Management
Tools
Aids in creating workplans
Identify all tasks, their sequence and estimate the

time to complete each one

Work breakdown structures (WBS): a hierarchy of
tasks to identify:
Duration of each task
Current status of each task
Task dependencies (shows which tasks must be

completed before others can begin)

Gantt charts: horizontal bar chart that shows the

WBS graphically
Network diagrams: PERT and CPM

Gantt Charts
Show tasks, duration, start and end and

resources (e.g. staff) in graphical format.

Activity
1 Study contextual aspects that contribute to language use and interpretation in English
2 Study use of spatial relations in Maori
3 Study contextual aspects that contribute to language use and interpretation in Maori
4 Compare English and Maori spatial language models and contextual aspects
3 Develop of Spatial Context Ontology
4 Develop of model for contextual language use
5 Develop of reasoningmethod
6 Evaluate models against test data and iteratively improve

Year 1

Year 2

Year 3

Scope Management
Scope creep
Occurs after the project is underway
Results from adding new requirements to the project
Can have a deleterious effect on the schedule

Techniques to manage the project scope:

Identify all requirements at the outset
Allow only those changes deemed absolutely necessary
Carefully examine the impact of suggested changes
Delay some changes for future enhancements
Time boxing

Environment & Infrastructure

Management
EnvironmentChoose the right set of tools
Use appropriate CASE tools to:
Increase productivity and centralize information (repository)
Utilize diagramsmore easily understood

Establish standards to reduce complexity

InfrastructureDocument the project appropriately

Store deliverables & communications in a project
binder
Use Unified Process standard documents
Dont put off documentation to the last minute

Case Study: Australian Capital

Territory SDMS
Goal: to automate, record, enable land

administration
Two existing systems:
CAD system
Oracle based textual system

Requirement: a single integrated system to

support all business functions

Case Study: Australian Capital

Territory SDMS
Waterfall model.
First analysed existing systems and

processes.
Designed new processes.
Designed user interfaces and applications to
support those processes.

Case Study: Australian Capital

Territory SDMS Lessons Learnt
Massive scope creep.
Moved to phased methodology.
User involvement much too late in the

process.
Full set of specifications: requirements,
functional, design.
Design spec impractical.
Very complex system technically and
organisationally.

What Next?