Different Approaches using Change Impact Analysis of UML Based Design for Software Development

IJISET - International Journal of Innovative Science, Engineering & Technology, Vol. 2 Issue 4, April 2015.
www.ijiset.com
ISSN 2348 – 7968
801
Different Approaches using Change Impact Analysis of
UML Based Design for Software Development
Ali Tariq Bhatti1
, Muhammad Murad Haider2
, Zill-e-Subhan2
1
North Carolina A&T State University, Greensboro NC USA
atbhatti@aggies.ncat.edu, ali_tariq302@hotmail.com
2
National University of Science and Technology (NUST), Islamabad Pakistan
murad.haider2003@gmail.com
2
National University of Computer and Emerging Sciences (FAST), Lahore Pakistan
zsubhan@hotmail.com
Abstract
Because of rapidly changing technologies,
requirements for the software systems are
constantly changing. This requires a change in
software design as well, as design should be
traceable to the requirements. There is a strong
need to deal with these changes in a systematic
manner. For this purpose, proper decision
making and change planning is required to
effectively implement the change. Change Impact
Analysis provides its services in this regard, by
allowing us to assess the potential side - effects
of change and also helps us in identifying that
what is needed to be modified to accomplish the
change. A number of impact analysis techniques
have been proposed that perform impact analysis
of UML based software design using a certain
strategy and methodology. In order to explore
the strengths and weaknesses of different
approaches toward impact analysis, this survey
paper includes an evaluation criterion for the
comparison of different impact analysis
techniques and a thorough analysis of these
techniques based on evaluation criteria.
1. Introduction.
As software systems become increasingly large
and complex, the need increases to predict and
control the effects of software changes. Software
can be modified to change the functionality of
the software and services and this change also
has wide effect on software design.The
perception that software can easily be changed is
often not the case, and understanding the
complete impact of a modification on the
software design is very important to managing
the propagation of changes. The Evolving nature
of software systems causes changes to the UML
models involved in the design of a software
project. So the impact of change must be
analyzed for ensuring the consistency, to make
these models up-to-date, and for assessing the
impact of change on overall system [4].
Management of change in software involves,
recognizing the need of change, evaluating the
impact of proposed change to the whole system,
adopting a technique for the accomplishment of
change, performing initial and consequent
modifications that are needed, and finally,
managing the versions to collect change related
data [7].
In case of a software design based on UML
models, the changes can in turn lead to
subsequent changes to other elements in the
UML diagrams. Impact analysis is defined as the
process of identifying the potential consequences
(side-effects) of a change, and estimating what
needs to be modified to accomplish that change
[3, 4].
Design changes have a strong impact on the
progress of a software development project and
can have adverse effects, if these changes are not
handled or analyzed in a systematic way. An
issue in this regard is that, there is a lack of
qualitative data, available to track the impact of
such design changes. These changes in many
cases increase the total development effort and
delay the project completion [6].

www.ijiset.com
ISSN 2348 – 7968
802
As changes in a software are introduced,
avoiding the defects becomes increasingly labor
intensive and error-prone. It exposes the systems
and organizations, the software serves, to risks
that are not visible without some impact analysis
technique to effectively estimate and accomplish
the change [1]. Whenever a change occurs in the
software design, many other inconsistencies may
be created in other design models, which are
directly or indirectly affected by the change.
Such propagation continues until change impact
analysis identifies all inconsistencies [2].
Inconsistencies may be automatically modeled
and detected by a set of consistency rules, also
called well-formed ness rules in OMG (2001)
[4].
In addition to consistency, dependency
relationship between UML models is also an
important factor to consider, achieving increased
flexibility and expressiveness. Dependence
relationship checking is also vital to achieve
correct impact analysis results because checking
of dependence relationship will help us to
identify the elements that are directly or
indirectly affected by the change.
Whenever a change occurs in UML based
software design, impact analysis must be
performed before implementation in order to
assess the potential side effects of the change
thus allowing early decision making and change
planning [4]. Impact analysis of the change is
necessary as it helps in estimating the cost of the
change, to understand the relationship between
the model, directly affected by the change, and
the structure of the software (dependence
relationship), and to determine the parts of the
software that are needed to be tested after the
implementation of change [5].
At design level, documentation of system’s
architecture is important as it aids in analyzing
the impact of change. The main advantage of
this documentation is its possible benefit as a
tool when making changes to the system in the
perspective of architectural evolution. The
rationale for the architecture is fundamental
information when analyzing how new or
changing requirements impact the design [9].
In [10], a three – part frame work is proposed,
which compares different impact analysis
approaches. These parts correspond to the
techniques to accomplish impact analysis and
effectiveness of each technique.
In order to evaluate different impact analysis
techniques, we have surveyed and analyzed
various approaches, focusing on the impact
analysis of changes in software design. The
strengths and weaknesses of different techniques
are assessed based on evaluation criteria,
presented in the paper.
The rest of the paper is structured as follows: the
succeeding section 2 contains discussion about
different impact analysis approaches, and an
evaluation criterion for comparison of different
impact analysis techniques. Section 3 contains a
detailed analysis of these approaches and section
4 outlines the main conclusions.
2. Impact Analysis Techniques.
Change impact analysis is about identifying the
possible side effects of a change and estimating
what needs to be modified to accomplish a
change. To perform the impact analysis different
authors have presented different approaches in
order to assess the impact of change on the
software development phases or artifacts.
The techniques that have yet been proposed to
support change impact analysis, mostly support
impact analysis at the code level of software
systems, but little effort has been made for
change impact analysis at the architectural or
design level.
Different approaches to impact analysis are
reported in literature that performs impact
analysis of change in the software at using
certain methodology. These techniques may vary
from each other, on how these techniques use to
accomplish impact analysis i.e. one technique
may rely only on checking the consistency of
UML diagrams , while another technique may
focus on checking the dependence relationship
among UML models along with consistency
checking. Following is the overview of some
approaches proposed by different authors:
2.1 Impact Analysis and Change
Management of UML Models (L.C.
Briand et al ,2003)
L. C. Briand et al [4] have proposed a UML
based approach to impact analysis. It involves
consistency checking of UML diagrams,
detection of change according to some change
taxonomy, determining with the help of formal
rules (defined by OCL or UML meta model), the
elements on which the changes have direct or

www.ijiset.com
ISSN 2348 – 7968
803
indirect impact; after these steps a distance
measure is computed to identify elements that
are more likely to require a change, for
prioritization of change impacts. This approach
also involves a prototype tool to help automate
the process of impact analysis.
2.2 Tracking the Impact of Design
Changes during Software Development
(Frank Padberg et al, 2001)
Padberg et al [6] have presented a methodology
to gather data about design changes in an
efficient way. It is proposed to use a “data base
of design changes”, consistently documenting
changes, as these occur in a software
development project. It will provide answers for
various critical issues like assessing the impact
of change and efforts for adjustments, this
methodology works for any kind of software
development project.
2.3 Automated impact analysis of UML
Models (L.C. Briand et al ,2005)
L.C. Briand et al [3] have presented a
methodology to perform impact analysis of UML
based software design, which is identical to the
one presented by L. C. Briand et al [4]. It
involves consistency checking of models, change
detection, determination of direct and indirect
impact of change on model elements and
calculation of distance measures. In addition two
case studies about Automated Teller Machine
(ATM) system and Cruise Control (CC) system
are also presented, in which impact analysis is
performed using prototype tool and at the same
time as, accuracy of proposed approach is
examined in another way (with respect to source
code).
2.4 Supporting Impact Analysis and
change Propagation in Software
Engineering Environments (Jun Han,
1997)
Jun Han [7] has presented an approach to address
the issues, concerning the change propagation
and impact analysis in the context of a generic
software engineering environment. The approach
emphasizes on three aspects of change
management that are; system representation,
impact analysis, and change propagation. The
main focus in this approach is on applying
change propagation and impact analysis using
original representations of software artifacts and
dependencies instead of extracted system
representations, using environment facilities like
specification of consistency property, artifact
manipulation, and defining relationships
(dependencies).
2.5 Software Change Impacts An
Evolving Perspective (Shawn A. Bohner,
2002)
Shawn A. bohner [1] presents an approach that
focus an on going research on impact analysis of
change in the systems that are increasingly using
middle-ware and integrated Commercial- Off-
The- Shelf (COTS) components. The focus in
this research is on extending the current software
change impact analysis technology (to support
middle-ware and COTS) in order to incorporate
interoperability dependency relationships and to
provide effective method for navigating software
changes using three dimensional (3D)
visualization techniques, to address the emerging
middle-ware problem.
2.6 Software Maintenance: An Approach
to Impact Analysis of Object Change
(Samuel Ajila, 1995)
Samuel Ajila [5] has proposed a technique in
which a tool is presented to assess the impact of
propagation of change (impact analysis). The
model presented is independent of any language
or design method and captures four software life
cycle phases which are requirement,
specification, design and coding. This technique
focuses on two kinds of dependencies i.e. inter –
phase dependency (dependency relation between
objects of different phases) and intra – phase
dependency (dependency relation between the
objects of the same phase).
2.7 An Object - Based, Attribute -
Oriented Approach for Software Change
Impact Analysis (Chung-Yang Chen et al,
2007)
Chung-Yang Chen et al [2] have proposed a
technique to accomplish change impact analysis
in software project management. The technique
employs an object – based, attribute oriented
technologies to determine the artifacts that are
affected by the change, and their relationships. It
also involves the implementation of a prototype
to support the automation of technique.
In order to compare different impact analysis
techniques parameters that are considered crucial
while comparing different techniques are Scope,
Nature, Prototype/ Tool support, Case study,
Consistency checking and Dependency Checking
In the following, an evaluation criterion is
presented for the analysis of aforementioned
impact analysis techniques on the bases of
parameters mentioned above:

www.ijiset.com
ISSN 2348 – 7968
804
Evaluation Parameter Description Possible Values
Scope Extent to which impact analysis is
performed.
Name of the artifact or software
engineering life cycle phase
methodology Whether the technique is code
based or model based
Code based, model based , both
or N/A (if it is not defined)
Prototype /Tool support Have the tool or prototype been
developed to support automation?
Yes/ No
Case study Whether the approach have been
applied to an example or
scenario.
Yes/No
Consistency Checking Whether the consistency of
artifacts is checked during CIA
Yes/ No
Dependency Checking Whether Dependency relationship
of artifacts is checked during
CIA.
Yes/ No
Table 1: Evaluation criteria for change impact analysis techniques.
3. Analysis.
Following is the detailed analysis of the
approaches presented by different authors, to
perform Software Change Impact Analysis
(CIA), based on critical evaluation parameters
presented in Table 1.
Table 2 present the detailed analysis of different
approaches related to software change impact
analysis according to the evaluation parameters
presented in table 1. Analysis shows that all the
approaches involve consistency checking except
the approaches presented by Frank Padberg et al
[6], Shawn. A. Bohner [1], Chung- Yang Chen et
al [2] and Samuel Ajila [5]. It is a limitation of
these approaches because consistency is crucial
in achieving correct results from impact analysis
algorithm.
An important issue here is to consider the
methodology employed by impact analysis
techniques i.e. whether a technique applies
model based or code based impact analysis
method or both. To perform impact analysis
using code based method, the change should be
implemented. Whereas model based methods
allow the accomplishment of impact analysis
before the implementation of change. So model
based methods are advantageous in the sense that
they ensure earlier decision making and change
planning. However, code based methods yield
more accurate results because the
implementation of the components is known.
From the detailed analysis performed in table 2,
it can also be found that all the approaches,
except Frank Padberg et al [6], Shawn. A.
Bohner [1], and Samuel Ajila [5], offers the
implementation of the approach on some case
study or real world examples which ensures the
applicability and practicality of these
approaches. So, there is a room for improvement
to ensure the practicality and to enhance the
understandability of their approaches, above
mentioned authors should also include the
application of their approaches on some case
studies or real world examples to validate both
the implementation and methodology.
All the approaches, except the approach
presented by Frank Padberg et al [6], perform
dependency relationship checking. It is also vital
to achieve correct impact analysis results.
Because checking of dependence relationship
will help us to identify the elements that are
directly or indirectly affected by the change, as
an element can have indirect impact of change if
it is dependent upon the element which is
directly affected by the change or in which the
change occurs.

www.ijiset.com
ISSN 2348 – 7968
805
All impact analysis techniques that are analyzed,
are automatable which provides room for
automated tool support for these approaches, in
order to automate the impact analysis process.
And also tools or prototypes for some of the
approaches have already been developed.
Table 2: Analysis of change impact analysis
approaches
Serial
#
Technique Scope Methodology Prototype/Tool
Support
Case
study
Consistency
checking
Dependency
Checking
1 L.C. Briand et al
(2003)
UML based
design models
Model based Yes Yes Yes Yes
2 L.C. Briand et al
(2005)
UML based
design models
Model based Yes Yes Yes Yes
3 Frank Padberg et
al
Software design N/A No No No No
4 Jun Han(1996) Software
development/main
tenance
Code based No Yes Yes Yes
5 Shawn
A.Bohner(2000)
CBSE COTs and
middle-ware
Code based No No No Yes
6 Chung-Yang Chen
et al (2007)
Software
development life
cycle
Both Yes Yes No Yes
7 Samuel Ajila
(1995)
Software
requirement,
specification,
design,
programming
Code based Yes No No Yes

www.ijiset.com
ISSN 2348 – 7968
806
4. Conclusion
Design changes occur frequently during software
development so the impact analysis of change in
software design has gained considerable
attention. With the growing use of UML models
for software design, the design of software
systems involves a large number of
interdependent UML diagrams. Because of
inevitability of change in software systems, these
diagrams also suffer from changes, to deal with
changes in the requirements, and which in turn
leads to subsequent changes in other dependent
elements in the UML diagrams. The introduction
of impact analysis of changes in software design
then adds more fidelity to change visibility. It
enables more accurate identification of side-
effects of change and identifies that what is
needed to be done in order to implement the
change. Various impact analysis techniques have
been proposed that perform impact analysis of
change in UML based software design using a
certain strategy and methodology (code based or
model based). In this survey paper, we have
performed a thorough survey of different impact
analysis techniques, on the basis of an evaluation
criterion which defines certain parameter that are
crucial to compare these different techniques.
Results of the analysis reveal that consistency
checking and dependency checking is not being
performed in certain impact analysis techniques
but this feature is very crucial to get accurate
impact analysis result. In addition, all impact
analysis techniques that are analyzed are
automatable that provide room for the
development of automated tool to support the
automation of these techniques. Furthermore
these impact analysis approaches have same
purpose but methodology employed by these
techniques, to perform impact analysis, is
different i.e. some approaches perform impact
analysis using code based methods while other
accomplish impact analysis using model based
methods.
References
[1] Bohner, S.A., “Software change impacts -an
evolving perspective”, Software
Maintenance, 2002. Proceedings. International
Conference on , vol., no., pp. 263-272, 2002
[2] Chung-Yang Chen; Cheung-Wo She; Jia-Da
Tang, “An object-based, attribute-oriented
approach for software change impact analysis”,
Industrial Engineering and Engineering
Management, 2007 IEEE International
Conference on , vol., no., pp.577-581, 2-4 Dec.
2007.
[3] Briand, L. C., Labiche, Y., O'Sullivan, L., and
Sówka, M. M. 2006. “Automated impact analysis
of UML models”. J. Syst. Softw. 79, 3 (Mar.
2006), 339-352. DOI=
http://dx.doi.org/10.1016/j.jss.2005.05.001
[4] Briand, L. C., Labiche, Y., and O'Sullivan, L.
2003. “Impact Analysis and Change Management
of UML Models”. In Proceedings of the
international Conference on Software
Maintenance (September 22 - 26, 2003). ICSM.
IEEE Computer Society, Washington, DC, 256.
[5] Ajila, S. 1995. “Software maintenance: an
approach to impact analysis of objects change”,
Softw. Pract. Exper. 25, 10 (Oct. 1995), 1155-
1181.DOI=http://dx.doi.org/10.1002/spe.4380251
006
[6] F. Padberg, “Tracking the impact of design
changes during software development”, EDSER-
3, May 2001, pp. 50-55.
[7] J. Han, “Supporting Impact Analysis and Change
Propagation in Software Engineering
Environments”, Proc. STEP’97 - 8th
International Workshop on Software Technology
and Engineering Practice, London (July 1997),
pp. 172-182.
[8] L. Briand, Y. Labiche and G. Soccar,
“Automating Impact Analysis and Regression
Test Selection Based on UML Designs” ,Carleton
University, Technical Report SCE-02- 04,
http://www.sce.carleton.ca/Squall/Articles/TR_S
CE-02-04.pdf, March, 2002, a short version
appeared in the proceedings of ICSM 2002
[9] Bratthall, L., Johansson, E., and Regnell, B.
2000. “Is a Design Rationale Vital when
Predicting Change Impact?” A Controlled
Experiment on Software Architecture Evolution.
In Proceedings of the Second international
Conference on Product Focused Software
Process Improvement (June 20 - 22, 2000). F.
Bomarius and M. Oivo, Eds. Lecture Notes In
Computer Science, vol. 1840. Springer-Verlag,
London, 126-139.
[10] Arnold, R. S. and Bohner, S. A. 1993. “Impact
Analysis - Towards a Framework for
Comparison”. In Proceedings of the Conference
on Software Maintenance D. N. Card, Ed. IEEE
Computer Society, Washington, DC, 292-301.

www.ijiset.com
ISSN 2348 – 7968
807
BIOGRAPHIES
Ali Tariq Bhatt received his Associate degree in
Information System Security (Highest Honors)
from Rockingham Community College, NC
USA, B.Sc. in Software engineering (Honors)
from UET Taxila, Pakistan, M.Sc in Electrical
engineering (Honors) from North Carolina A&T
State University, NC USA in 2010, and currently
pursuing PhD in Electrical engineering from
North Carolina A&T State University. His
current research interests include Coding
Algorithm, Networking Security, Mobile
Telecommunication, Biosensors, Genetic
Algorithm, Swarm Algorithm, Health
Bioinformatics, Control system, Power, Software
development, Communication, and Signal
Processing.

Different Approaches using Change Impact Analysis of UML Based Design for Software Development

More Related Content

What's hot (19)

Similar to Different Approaches using Change Impact Analysis of UML Based Design for Software Development (20)

More from zillesubhan (8)

Recently uploaded (20)

Different Approaches using Change Impact Analysis of UML Based Design for Software Development