See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233662416

A fuzzy multi-criteria decision approach for software development strategy

selection

Article  in  International Journal of General Systems · April 2004

DOI: 10.1080/03081070310001633581

CITATIONS READS
201 346

3 authors, including:

Cengiz Kahraman
Istanbul Technical University
686 PUBLICATIONS   16,095 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

reliability measurement and improvement View project

Ship Design Project View project

All content following this page was uploaded by Cengiz Kahraman on 14 November 2014.

The user has requested enhancement of the downloaded file.

This article was downloaded by: [Istanbul Technical University]
On: 14 November 2014, At: 07:47
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

International Journal of General Systems

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/ggen20

A fuzzy multi-criteria decision approach for software

development strategy selection
a b c
Gülçin Büyüközkan , Cengiz Kahraman & Da Ruan
a
Galatasaray University, Department of Industrial Engineering , Ortaköy, 34357, Istanbul,
Turkey
b
Department of Industrial Engineering , Istanbul Technical University , Macka, 34367,
Istanbul, Turkey
c
Belgian Nuclear Research Centre (SCK•CEN) , Boeretang 200, B-2400, Mol, Belgium
Published online: 26 Jan 2007.

To cite this article: Gülçin Büyüközkan , Cengiz Kahraman & Da Ruan (2004) A fuzzy multi-criteria decision approach
for software development strategy selection, International Journal of General Systems, 33:2-3, 259-280, DOI:
10.1080/03081070310001633581

To link to this article: http://dx.doi.org/10.1080/03081070310001633581

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 International Journal of General Systems, April–June 2004 Vol. 33 (2–3), pp. 259–280

A FUZZY MULTI-CRITERIA DECISION APPROACH

FOR SOFTWARE DEVELOPMENT STRATEGY
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

SELECTION
GÜLÇIN BÜYÜKÖZKANa,*, CENGIZ KAHRAMANb and DA RUANc
a
Galatasaray University, Department of Industrial Engineering, Ortaköy 34357, Istanbul, Turkey;
b
Istanbul Technical University, Department of Industrial Engineering, Macka 34367, Istanbul, Turkey;
c
Belgian Nuclear Research Centre (SCK†CEN), Boeretang 200, B-2400 Mol, Belgium

(Received 24 February 2003; In final form 18 August 2003)

This study proposes a methodology to improve the quality of decision-making in the software development project
under uncertain conditions. To deal with the uncertainty and vagueness from subjective perception and experience of
humans in the decision process, a methodology based on the extent fuzzy analytic hierarchy process modeling to
assess the adequate economic (tangible) and quality (intangible) balance is applied. Two key factors of economic and
quality are evaluated separately by fuzzy approaches and both factors’ estimates are combined to obtain the
preference degree associated with each software development project strategy alternative for selecting the most
appropriate one. Using the proposed approach, the ambiguities involved in the assessment data can be effectively
represented and processed to assure a more convincing and effective decision-making. Finally, a real case-study is
given to demonstrate the potential of the methodology.

Keywords: Fuzzy multi-criteria decision-making, fuzzy analytic hierarchy process, software cash flows, software
quality

1. INTRODUCTION

Today’s markets are generally perceived to be demanding much higher quality and higher
performing products at a reasonably low cost. Software development is a special case for
product development. There exist lots of differences that are worthwhile to notice between
software production and traditional production. It is not repetitive, individual attributes of the
personnel that have an extreme influence on the quality of the product, the failure probability
in one production step is highly dependent on some previous steps. This would imply that
uncertainty factors dominate the whole production process.
Despite some extensive research in software development (Glass, 1998; Schenk et al.,
1998; Verner et al., 1999), many software development projects are still failing (McConnell,
1996). In 1994, 31% of all corporate software development projects resulted in cancellation
(Glass, 1998). As shown in (Hoffman, 1999), failure rates for software development projects

*Corresponding author. Fax: þ 90-212-259-5557. E-mail: gbuyukozkan@gsu.edu.tr

ISSN 0308-1079 print/ISSN 1563-5104 online q 2004 Taylor & Francis Ltd
DOI: 10.1080/03081070310001633581
 260 G. BÜYÜKÖZKAN et al.

are as high as 85%. This situation forces the decision-makers to become more
attentive through better decision-making for all the phases of the software development
process.
To obtain the best performance from software development, the efficient and
effective management of this process is vital. The decision-maker must take into
account the customers’ needs, the company’s strategies as well as technological
opportunities and the company’s resources, and deduce the goals based on these factors
for a successful software development. One of the most critical decisions while managing a
development project is the project strategy selection. Several researchers have suggested that
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

it is difficult for managers to end up with projects once they start. For this reason, the current
paper focuses especially on making efficient software development and related strategy
decisions.
The analytic hierarchy process (AHP) (Saaty, 1980) is one of the extensively used multi-
criteria decision-making analysis tools for modeling unstructured problems in different areas
such as political, economic, social, and management sciences. The AHP assumes that the
multi-criteria problem can be completely expressed in a hierarchical structure. The data
acquired from the decision-makers are pairwise comparisons concerning the relative
importance of each of the criteria, or the degree of preference of one factor to another with
respect to each criterion. Since it is difficult to map qualitative preferences to point estimates,
a degree of uncertainty will be associated with some or all-pairwise comparison values in an
AHP problem. The problem of generating such a priority vector in the uncertain pair-to-pair
comparison environment is called the fuzzy AHP problem.
In order to deal with the uncertainty and vagueness from subjective perception and
experience of humans in the decision process, a methodology based on Chang’s (1996) extent
fuzzy AHP modeling to assess the adequate economic (tangible) and quality (intangible)
balance is proposed. Economic and quality factors are evaluated separately and both factors’
estimates are combined to obtain the preference degree associated with each software
development project and its strategy alternatives for selecting the most appropriate one.
The ambiguities involved in the assessment data can be effectively represented and processed
by the proposed approach to assure a more convincing and effective decision-making.
The rest of the paper is organized as follows. The second section describes briefly the
fuzzy AHP methodology. The third section discusses the proposed software development and
its strategy selection approach. The fourth section presents a case study from a Turkish
software developer company. Finally, the last section contains some concluding remarks and
perspectives.

2. THE FUZZY AHP APPROACH

The proposed software strategy evaluation methodology is based mostly on the fuzzy AHP
method. For this reason, this section describes this methodology. In the following, the
literature review on fuzzy AHP is firstly given and then the extent analysis method on fuzzy
AHP that will be used for the proposed evaluation methodology is outlined.

2.1 Literature Review for Fuzzy AHP

There are many fuzzy AHP methods proposed by various authors. These methods are
systematic approaches to an alternative selection and justification problem by using the
concept of fuzzy set theory (Zadeh, 1965) and hierarchical structure analysis. Decision-
makers usually find that it is more confident to give interval judgments than fixed value
 A FUZZY MULTI-CRITERIA DECISION 261

judgments. This is because usually he/she is unable to be explicit about his/her

preferences due to the fuzzy nature of the comparison process.
The earliest work in fuzzy AHP appeared in Van Laarhoven and Pedrycz (1983),
which compared fuzzy ratios described by triangular membership functions. Buckley
(1985) determined fuzzy priorities of comparison ratios with trapezoidal membership
functions. Stam et al. (1996) explored how recently developed artificial intelligence
techniques can be used to determine or approximate the preference ratings in AHP.
They concluded that the feed-forward neural network formulation seemed to be a
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

powerful tool for analyzing discrete alternative multi-criteria decision problems with
imprecise or fuzzy ratio-scale preference judgments. Chang (1996) introduced an
approach for handling fuzzy AHP, with the use of triangular fuzzy numbers for pairwise
comparison scale of fuzzy AHP, and the use of the extent analysis method for the
synthetic extent values of the pairwise comparisons. Cheng (1996) proposed another
algorithm for evaluating naval tactical missile systems by the fuzzy analytical hierarchy
process based on grade value of membership function. Weck et al. (1997) presented a
method to evaluate different production cycle alternatives adding the mathematics of
fuzzy logic to the classical AHP. Any production cycle evaluated in this manner yields a
fuzzy set. The outcome of the analysis can finally be defuzzified by forming the surface
center of gravity of any fuzzy set, and the alternative production cycles investigated can
be ranked in order in terms of the main objective set. Deng (1999) presented a fuzzy
approach for tackling qualitative multi-criteria analysis problems in a simple and
straightforward manner. Lee et al. (1999) reviewed basic ideas behind the AHP. Based
on these ideas, they introduce the concept of a comparison interval and propose a
methodology based on stochastic optimization to achieve global consistency and to
accommodate the fuzzy nature of the comparison process. Cheng et al. (1999) proposed
a new method for evaluating weapon systems by an analytical hierarchy process based
on linguistic variable weight. Zhu et al. (1999) discussed some extent analysis methods
and applications of fuzzy AHP. Leung and Cao (2000) proposed a fuzzy consistency
definition with consideration of a tolerance deviation for alternatives in fuzzy AHP.
Essentially, the fuzzy ratios of relative importance, allowing certain tolerance deviation,
are formulated as constraints on the membership values of the local priorities. The fuzzy
local and global weights are determined via the extension principle. The alternatives are
ranked on the basis of the global weights by application of the maximum – minimum set
ranking method. Kuo et al. (2002) developed a decision support system for locating a
new convenience store. The first component of the proposed system is the hierarchical
structure development for fuzzy AHP. More recently, Yu (2002) incorporated an
absolute term linearization technique and a fuzzy rating expression into a goal
programming (GP)-AHP model for solving group decision-making fuzzy AHP problems
by employing the property of GP to treat a fuzzy AHP problem. The most recent work
related to this subject would be the research by Wang and Lin (2003) on a fuzzy multi-
criteria group decision-making approach to select configuration for software
development.
Table I gives the comparison of the fuzzy AHP methods in the literature, which have
important differences in their theoretical structures. The comparison includes the advantages
and disadvantages of each method. In this paper, we prefer Chang’s (1992; 1996) extent
analysis method since the steps of this approach are relatively easier than the other fuzzy
AHP approaches and similar to the conventional AHP. We also integrate the improvement
proposed by Zhu et al. (1999) as presented in the following part. Recently, Kwong and Bai
(2003) applied this method to prioritize customer requirements in the QFD while Bozdağ
et al. (2003) used this approach in the evaluation of CIM alternatives.
 262 G. BÜYÜKÖZKAN et al.

TABLE I The comparison of different fuzzy AHP methods

Sources The main characteristics of the method Advantages (A) and disadvantages (D)
Van Laarhoven and † Direct extension of Saaty’s (A) The opinions of multiple decision-
Pedrycz (1983) AHP method with triangular makers can be modeled in the
fuzzy numbers reciprocal matrix
† Lootsma’s logarithmic least (D) There is not always a solution to
square method is used to derive the linear equations
fuzzy weights and fuzzy
performance scores
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

(D) The computational requirement is

tremendous, even for a small
problem
(D) It allows only triangular fuzzy
numbers to be used

Buckley (1985) † Extension of Saaty’s AHP (A) It is easy to extend

method with trapezoidal to the fuzzy case
fuzzy numbers
† Uses the geometric mean method (A) It guarantees a unique
to derive fuzzy weights and solution to the reciprocal
performance scores comparison matrix
(D) The computational requirement
is tremendous

Boender et al. (1989) † Modifies van Laarhoven and (A) The opinions of multiple decision-
Pedrycz’s method makers can be modeled
† Presents a more robust (D) The computational requirement
approach to the normalization is tremendous
of the local priorities

Chang (1996) † Synthetical degree values (A) The computational requirement

is relatively low
† Layer simple sequencing (A) It follows the steps of crisp AHP.
It does not involve additional
operations
† Composite total sequencing (D) It allows only triangular fuzzy
numbers to be used

Cheng (1996) † Builds fuzzy standards (A) The computational requirement is

† Represents performance scores by
membership functions
† Uses entropy concepts to calculate
aggregate weights
not tremendous
(D) Entropy is used when probability
distribution is known. The method
is based on both probability and
possibility measures

2.2 Extent Analysis Method on Fuzzy AHP

Let X ¼ {x1 ; x2 ; . . .; xn } be an object set, and U ¼ {u1 ; u2 ; . . .; um } be a goal set. According to
the method of Chang’s extent analysis model, each object is taken and extent analysis for
each goal, gi is performed, respectively (Chang, 1992; 1996). Therefore, m extent analysis
values for each object can be obtained, with the following signs:

M 1gi ; M 2gi ; . . .; M m
gi ; i ¼ 1; 2; . . .; n ð1Þ

where all the M jgi ð j ¼ 1; 2; . . .; mÞ are triangular fuzzy numbers (TFNs).

 A FUZZY MULTI-CRITERIA DECISION 263

The steps of the improved Chang’s extent analysis model (Zhu et al., 1999), which
we apply in this study, can be given as follows.
Step 1. The value of fuzzy synthetic extent with respect to the i th object is defined as
" #21
Xm Xn Xm
j j
Si ¼ M gi ^ M gi : ð2Þ
j¼1 i¼1 j¼1
Pm j
To obtain j¼i M gi ,
perform the fuzzy addition operation of m extent analysis values for
a particular matrix such that
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

!
X
m X m X
m X
m
j
M gi ¼ lj ; mj ; uj ð3Þ
j¼i j¼1 j¼1 j¼1
hP Pm i21
n j
and to obtain i¼1 j¼1 M gi ; perform the fuzzy addition operation of
j
M gi ð j ¼ 1; 2; . . .; mÞ values such that
!
X n X m Xn X
n X
n
j
M gi ¼ li ; mi ; ui ð4Þ
i¼1 j¼1 i¼1 i¼1 i¼1

and then compute the inverse of the vector in Eq. (4) such that
0 1
" #21 1 1 1
Xn X m ; ;
BX n X
n X C
n
M jgi ¼@ ui mi li A: ð5Þ
i¼1 j¼1 i¼1 i¼1 i¼1

Step 2. The degree of possibility of M2 $ M1 is defined as

VðM 2 $ M 1 Þ ¼ sup bminðmM 1 ðxÞ; mM 2 ðyÞc: ð6Þ
y$x

When a pair (x, y) exists such that y $ x and mM 1 ðxÞ ¼ mM 2 ðyÞ; then we have VðM 2 $ M 1 Þ ¼ 1:
Since M 1 ¼ ðl1 ; m1 ; u1 Þ and M 2 ¼ ðl2 ; m2 ; u2 Þ are convex fuzzy numbers we have that
8
> 1; if m2 $ m1
>
<
VðM 2 $ M 1 Þ ¼ hgtðM 1 > M 2 Þ ¼ mM 2 ðdÞ ¼ 0; if l1 $ u2 ð7Þ
>
> l 2u
: ðm 2u Þ2ðm 2l Þ ; otherwise;
1 2
2 2 1 1

where d is the ordinate of the highest intersection point D between mM 1 and mM2 (see Fig. 1).
To compare M1 and M2, we need both the values of VðM 1 $ M 2 Þ and VðM 2 $ M 1 Þ:
Step 3. The degree possibility for a convex fuzzy number to be greater than k convex fuzzy
numbers M i ði ¼ 1; 2; . . .; kÞ can be defined by
VðM $ M 1 ; M 2 ; . . .; M k ¼ V½ðM $ M 1 Þ and ðM $ M 2 Þ and. . .and ðM $ M k Þ
¼ min VðM $ M i Þ; i ¼ 1; 2; 3; . . .; k: ð8Þ

Assume that
d0 ðAi Þ ¼ min VðSi $ Sk Þ: ð9Þ
For k ¼ 1; 2; . . .; n; k – i: Then the weight vector is given by
W 0 ¼ ðd0 ðA1 Þ; d0 ðA02 Þ; . . .; d0 ðAn ÞÞT ð10Þ
where Ai ði ¼ 1; 2; . . .; nÞ are n elements.
 264 G. BÜYÜKÖZKAN et al.
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

~ 1 and M
FIGURE 1 The intersection between M ~ 2:

Step 4. Via normalization, the normalized weight vectors are

W ¼ ðdðA1 Þ; dðA2 Þ; . . .; dðAn ÞÞT ; ð11Þ
where W is not a fuzzy number.

3. SOFTWARE DEVELOPMENT STRATEGY SELECTION METHODOLOGY

3.1 Evaluation Methodology and its Hierarchical Structure

Software is a computer program with all associated documentation and configuration data,
which is needed to make the program operate correctly. Software products may be developed
for a particular customer (customized products) or for a general market (generic products).
In this study, we propose to analyze the project strategy selection of a software developer
company, which regularly produces different generic software for the market.
The starting point of our study is to use the right software development strategy by having
the right balance of software cost and software quality. However, software development
investments have some special characteristics that make it very complicated to evaluate their
costs and quality. First, they are intangible in nature. In many cases monetary measures
cannot be used, but subjective arguments have to be applied also. Second, diverse interest
groups recognize the quality factors and their importance differently. Third, these software
development investments are irreversible in nature. Therefore an appropriate evaluation
methodology and evaluation criteria have to be identified while keeping these special
characteristics of the software development project and the requirements of the company.
The procedure of proposed decision methodology is divided briefly into the four main steps
as summarized in Table II.
The first phase of the fuzzy AHP modeling process involves formulating an appropriate
hierarchy of the fuzzy AHP model consisting of the goal, strategic factors, criteria, and the
alternatives. The goal of our problem is to select the appropriate development strategy that
can meet customer requirements, bring profits to the company, and compete strongly in the
market. This goal is placed on the first level of the hierarchy as shown in Fig. 2. Two strategic
factors, namely economic and quality are identified through the related references and
 A FUZZY MULTI-CRITERIA DECISION 265

TABLE II The steps of software development strategy selection methodology

Step 1 Define the evaluation factor and criteria for software development project strategy selection and
construct the hierarchical structure development of the fuzzy AHP
Step 2 Calculate the relative importance of economic factor and quality factor and quality criteria
according to “software economic evaluation model” or “software quality evaluation model”
Step 3 Aggregate the relative importance of alternatives versus economic and quality factors with fuzzy
AHP methodology
Step 4 Compute the overall score of each software development strategy and make a decision for a
further development
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

interview of software experts to achieve this goal, which form the second level of the
hierarchy. The economic factor is important because the lower the cost of a development, the
higher the productivity and efficiency, thus bringing more profit to the company. Quality is
equally important as it focuses more on meeting customers’ requirements and becoming
competitive in order to stay ahead in the marketplace.
The third level of the hierarchy occupies the criteria defining the strategic factor of quality
of the second level. There are three criteria related to software quality, namely design,
performance and adaptation, which will be detailed in the following parts. The fourth level
consists of the different alternatives for evaluation. Fig. 2 gives the software development
project strategy selection hierarchical structure that is used in this study.

3.2 A Fuzzy Model to Evaluate Software Development Cash Flows

In fuzzy AHP, decision-makers usually attain the attribute weights using their experiences.
Our goal is to base the attribute weights on monetary data. First, the fuzzy present worth
(PW) of the cash flows of each alternative is calculated. Then, the extent analysis method on
fuzzy AHP is used to select the best alternative. While calculating the fuzzy PW, the least
common multiple of lives for the alternatives with different lives will not be used since the
software alternatives are not repeatable. At this point we firstly summarize the fuzzy PW
method before explaining the steps of the proposed fuzzy economic model.

FIGURE 2 Software development project strategy selection hierarchical structure.

 266 G. BÜYÜKÖZKAN et al.

3.2.1 Fuzzy Present Worth (PW) Method

The PW method of alternative evaluation is very popular because future expenditures or
receipts are transformed into equivalent dollars now. That is, all of the future cash flows
associated with an alternative are converted into present dollars. If the alternatives have
different lives, the alternatives must be compared over the same number of years.
Quite often in finance, future cash amounts and interest rates are estimated. One usually
employs educated guesses, based on expected values or other statistical techniques, to obtain
future cash flows and interest rates. A statement like “approximately between 10 and 15%”
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

must be translated into an exact amount such as “12.5%”. Appropriate fuzzy numbers can be
used to capture the vagueness of “approximately between 10 and 15%.”
Chiu and Park (1994) proposed a PW formulation of a fuzzy cash flow. The result of the
PW is also a fuzzy number with a non-linear membership function. The PW can be
approximated by a TFN. Chiu and Park’s (1994) formulation is
2 0     1
X
n max Plðt yÞ ; 0 min Ptlð yÞ ; 0
~ ¼4
PW @
Qt  þQ   A;
rð yÞ t lð yÞ
t¼0 t0 ¼0 1 þ r t0 t0 ¼0 1 þ r t0

0     13
X
n max Prð yÞ
t ;0 min Prð yÞ
t ;0
@  A5
Qt  lð yÞ
þQ 
t rð yÞ
t¼0 0
t ¼0 1 þ r t 0 0
t ¼0 1 þ r t 0 ð12Þ

where Ptlð yÞ : the left representation of the cash at time t, Prð

t
yÞ
: the right representation of the
cash at time t, r t : the left representation of the interest rate at time t, r rðyÞ
lðyÞ
t : the right
representation of the interest rate at time t.
Buckley’s (1987) membership function for P~ n ;
 
mðxjP~ n Þ ¼ pn1 ; fn1 ð yjP~ n Þ=pn2 ; pn2 =f n2 ð yjP~ n Þ; pn3 ð13Þ

is determined by
fni ð yjP~n Þ ¼ f i ð yjFÞð1
~ þ f k ð yj~r Þ2n ð14Þ
for i ¼ 1; 2 where k ¼ i for negative F~ and k ¼ 3 2 i for positive F:
~
Ward (1989) gave the fuzzy PW function as

~ ¼ ð1 þ rÞ2n ða; b; c; dÞ
PW ð15Þ

where (a, b, c, d) is a flat fuzzy filter function (4Fs) number.

For the uniform cash flow series, PW is calculated by

ð1 þ rÞn 2 1
PW ¼ Ag ðn; rÞ ¼ A : ð16Þ
ð1 þ rÞn r

In the case of fuzziness, the membership function mðxjPW ~ n Þ for PW

~ n is determined by
     
~ n ¼ f i yjA~ g n; f 32i ð yj~r Þ
f ni yjPW ð17Þ
~ is
and TFN( y) for fuzzy PW
   
~ n ð yÞ ¼ AlðyÞ g ðn; rrðyÞ ; ArðyÞ g n; r lðyÞ Þ :
PW ð18Þ
 A FUZZY MULTI-CRITERIA DECISION 267

To compare the fuzzy PWs of alternative projects, it is necessary to use a ranking method
to rank TFNs. There are many ranking methods and they may give different ranking results
and most methods are tedious in graphic manipulation requiring complex mathematical
calculation. Liou and Wang (1992) proposed the total integral value method with an index of
optimism v [ [0,1]. Let A~ be a fuzzy number with left membership function f LA~ and right
membership function f RA~ : Then the total integral value is defined as:

~ ¼ vER ðAÞ
Ev ðAÞ ~ þ ð1 2 vÞEL ðAÞ
~ ð19Þ
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

where
ðb
~ ¼
ER ðAÞ xf RA~ ðxÞ dx ð20Þ
a
ðd
~ ¼
EL ðAÞ xf LA~ ðxÞ dx ð21Þ
g

where 21 a a # b # g # d a þ1 and a trapezoidal fuzzy number is denoted by

(a,b,g,d). For a TFN, A~ ¼ ða; b; cÞ;

~ ¼ 1 ½vða þ bÞ þ ð1 2 vÞðb þ cÞ

Ev ðAÞ ð22Þ
2
and for a trapezoidal fuzzy number, B~ ¼ ða; b; g; dÞ;

~ ¼ 1 
Ev ðBÞ vðg þ dÞ þ ð1 2 vÞða þ bÞ : ð23Þ
2

3.2.2 The Steps of the Proposed Model

The capital investment (CI) levels in the software sector are classified by using TFNs. The
TFNs are named as “very low (VL),” “low (L),” “normal (N),” “high (H),” and “very high
(VH).” The operating expenditure (OE) levels in the software sector are also classified by
using the same classification.
Now assume that a firm in the software sector will invest in one of possible alternative
projects. The steps of the model are given in the following.

1. Ask for the possible capital investment level for each alternative: around what amount
will the capital investment level be? If the decision-maker does not prefer giving
a numerical value, ask the following question: will the CI level be “very low,” “low,”
“normal,” “high,” or “very high?” (see Fig. 3).
2. Ask for the mean level of possible OE per year in the future: around what amount will the
OE per year be? If the decision-maker does not prefer giving a numerical value, ask the
following question: will the OE level be “very low,” “low,” “normal,” “high,” or “very
high?” (see Fig. 4).
3. Ask for the mean level of possible monthly benefits (B) from services after sale: around
what amount will B be? If the decision-maker does not prefer giving a numerical value,
ask the following question: Will the B level be “very low,” “low,” “normal,” “high,” or
“very high?” (see Fig. 5).
4. Determine the crisp planning horizon and the fuzzy minimum attractive rate of return.
5. Compute the fuzzy net PW which is the difference between fuzzy PWs of benefits and
costs.
 268 G. BÜYÜKÖZKAN et al.
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

FIGURE 3 Capital investment levels.

FIGURE 4 Operating expenditure levels.

FIGURE 5 Monthly benefit levels.

3.3 A Fuzzy Model to Evaluate Software Quality

Software economic elements are quantitative, and therefore can be measured and evaluated
more easily. Software quality, on the other hand, is a multi-dimensional concept, difficult to
define and measure (Ashrafi, 2002). Two levels of quality attributes are distinguished: quality
criteria and attributes. The former cannot be measured directly, while the latter can be
measured subjectively (Hines and Goerner, 1995). Based on this, there exist different software
quality standards and frameworks such as ISO 9001/9000-3, extended ISO model (Zeist and
Hendriks, 1996), the capability maturity model (Paulk, 1995) require or recommend software
quality measurements (Jørgensen, 1999; Hareton and Leung, 2001). In this study, we used the
software quality model proposed by the Software Quality Assurance (SQA) group. This model
identifies 14 quality attributes in three criteria, as shown in Table III.
 A FUZZY MULTI-CRITERIA DECISION 269

TABLE III Software quality model (Source: The Handbook of Software Quality Assurance, Prentice Hall, 1998)

Criteria Attributes
Software design Correctness: extent to which the software conforms to its specifications and
quality conforms to its declared objectives
Maintainability: ease of effort for locating and fixing a software failure within
a specified time period
Verifiability: ease of effort to verify software features and performance based
on its stated objectives
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

Software performance Efficiency: extent to which the software is able to do more with less system
quality (hardware, operating system, communications, etc.) resources
Integrity: extent to which the software is able to withstand intrusion by
unauthorized users or software within a specified time period
Reliability: extent to which the software will perform (according to its stated
objectives) within a specified time period
Usability: relative ease of learning and the operation of the software
Testability: ease of testing program to verify that it performs a specified function

Software adaptation Expandability: relative effort required to expand software capabilities and/or
quality performance by enhancing current functions or by adding new functionality
Flexibility: ease of effort for changing the software’s mission, functions or
data to meet changing needs and requirements
Portability: ease of effort to transport software to another environment
and/or platform
Reusability: ease of effort to use the software (or its components) in another software

Throughout this paper, it is assumed that the decision-makers utilize the linguistic
weighting set W, W ¼ {ALI; VSLI; SLI; WLI; EI; WMI; SMI; VSMI; AMI}, where ALI
is absolutely less important; VSLI, very strongly less important; SLI, strongly less
important; WLI, weakly less important; EI, equally important; WMI, weakly more
important; SMI, strongly more important; VSMI, very strongly more important and AMI
is absolutely more important, to evaluate the importance of the software quality attributes
and criteria. The triangular fuzzy conversion scale given in Table IV is used in our
evaluation model.
The important weights of decision criteria and attributes are obtained using the Delphi
method with eight experts on software development. In general, the Delphi study is a
method for structuring a group communication process so that the process is effective in
allowing individuals to deal with complex problems (Delbecq et al., 1975). The method
consists principally of knowledgeable and expert contributors individually completing a
form and submitting the results to a central coordinator. The coordinator processes the
contributions, looking for central and extreme tendencies, and the rationales therefor.

TABLE IV Triangular fuzzy conversion scale

Linguistic scale Triangular fuzzy scale Triangular fuzzy reciprocal scale

Just equal (1, 1, 1) (1, 1, 1)
Equally important (1/2, 1, 3/2) (2/3, 1, 2)
Weakly more important (1, 3/2, 2) (1/2, 2/3, 1)
Strongly more important (3/2, 2, 5/2) (2/5, 1/2, 2/3)
Very strongly more important (2, 5/2, 3) (1/3, 2/5, 1/2)
Absolutely more important (5/2, 3, 7/2) (2/7, 1/3, 2/5)
 270 G. BÜYÜKÖZKAN et al.

The results are then fed back to the respondent group, who are asked to resubmit their
views, assisted by the “new” input provided by the coordinator. The Delphi method
was successfully used in technical and business-related evaluation systems and has a
methodical advantage compared to other group discussion methods due to anonymity of
experts and avoidance of the dominance of singular opinions (Kenis, 1995).
The application of the quality evaluation model will be given in the following section
throughout the case study.
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

4. APPLICATION OF THE FUZZY AHP MODEL TO A SPECIFIC TURKISH

SOFTWARE DEVELOPER COMPANY

A Turkish software company, YAZGEL, is a specialist to develop the generic management

software for the Turkish SMEs. It is decided to develop software that will support the project
managers, and there are three development strategies (namely A, B, and C). The managers
want to make the best selection among them.

4.1 Calculations Using the Economic Model

The development team of YAZGEL is uncertain about the exact monetary levels of the
investments. The levels per month of capital investment, operating expenditures and
benefits in the software sector in Turkey are given in Figs. 6, 7 and 8, respectively.
Table V summarizes the data given by the managers of YAZGEL.
The interest rate, r, will be around 2% per month. It will be represented by the TFN, (1.5,
2.0 and 2.5%) per month:
r~ ¼ ð1:5%; 2:0%; 2:5%Þ ¼ ð0:015 þ 0:005y; 0:025 2 0:005yÞ ¼ ð f 1 ð yj~r Þ; f 2 ð yj~r ÞÞ:

4.1.1 Calculation of the Present Worth of Alternative A

The cash flow diagram for A is shown in Fig. 9.
Considering Figs. 6– 8:
f 1 ð yjCI~A Þ ¼ 12; 000 þ 3000y;
f 1 ð yjB~ A Þ ¼ 2000 þ 1000y;
f 1 ð yjOE~ A Þ ¼ 250 þ 250y;
f 2 ð yjC I~A Þ ¼ 18; 000 2 3000y
f 2 ð yjB~ A Þ ¼ 4000 2 1000y
f 2 ð yjOE~ A Þ ¼ 750 2 250y:

FIGURE 6 Capital investment levels in the software sector in Turkey.

 A FUZZY MULTI-CRITERIA DECISION 271
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

FIGURE 7 Operating expenditure levels in the software sector in Turkey.

FIGURE 8 Benefits in the software sector in Turkey.

The fuzzy present worth of capital investments, FPW 2 CI~A ; is calculated as follows
(Kahraman et al., 2000; 2002; Kahraman, 2001):

0 12;000 þ 3000y 12;000 þ 3000y 12;000 þ 3000y

1
1:02520:005y þ ð1:02520:005yÞ2
þ ð1:02520:005yÞ 3 ;
B C
FPW 2 C I~ A ¼ @ 18;00023000y 18;00023000y 18;00023000y A
1:015 þ 0:005y þ ð1:015 þ 0:005yÞ2
þ ð1:015 þ 0:005yÞ3

¼ ð34; 272:283; 43; 258:249; 52; 419:607Þ:

TABLE V The data of the alternatives

Capital Operating Benefits from Development Useful

investment expenditure service after periods lives
Alternatives per month per month sale per month (month) (year)
A Around $15,000 Around $500 Around $3000 3 2
B Around $12,000 Around $750 Around $4000 5 3
C Around $18,000 Around $1000 Around $5000 5 4
 272 G. BÜYÜKÖZKAN et al.

FIGURE 9 Cash flow diagram for alternative A.

Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

FIGURE 10 Cash flow diagram for alternative B.

The fuzzy net cash flow per month, NC F~ A ; is

NC F~ A ¼ B~ A 2 OE~ A ¼ ð f 1 ð yjNC F~ A Þ; f 2 ð yjNC F~ A ÞÞ ¼ ð1250y þ 1250; 3750 2 1250yÞ:

The fuzzy PW of, NC F~ A ; FPW 2 NC F~ A ; is

0 h i 1
2 0:005yÞ24 2 1
ð1250y þ 1250Þ ð1:025 ð1:025 ;
2 0:005yÞ24 ð0:025 2 0:005yÞ C
B
~
FPW 2 NC F A ¼ @ B h i C
ð1:015 þ 0:005yÞ24 21 A
ð3750 2 1250yÞ ð1:015 þ 24
0:005yÞ ð0:015 þ 0:005yÞ

¼ ð22; 356:232; 47; 284:814; 75; 114:02Þ:

~ is
The fuzzy net present worth of the net cash flow at point 0, FNPW;
~ A ¼ ðFPW 2 NC F~ A Þi ð1 þ f 32i ð yj~r ÞÞ23 2 ðFPW 2 C I~A Þ
FNPW

22; 356:232 þ 24; 928y 75; 114:02 2 27; 829:206y
¼ ; 2 ðFPW 2 C I~ A Þ
ð1:025 2 0:005yÞ3 ð1:015 þ 0:005yÞ3

¼ $ð231; 659:623; 1298:739; 37; 560:531Þ:

4.1.2 Calculation of the Present Worth of Alternatives B and C

The cash flow diagram for B is given in Fig. 10.
The cash flow diagram for C is given in Fig. 11.

FIGURE 11 Cash flow diagram for alternative C.

 A FUZZY MULTI-CRITERIA DECISION 273

TABLE VI Evaluation matrix of software development strategies with respect to the software economic

A B C
A (1, 1, 1) (1/3, 2/5, 1/2) (2/7, 1/3, 2/5)
B (2, 5/2, 3) (1, 1, 1) (2/5, 1/2, 2/3)
C (5/2, 3, 7/2) (3/2, 2, 5/2) (1, 1, 1)

~ B and FNPW
Using the similar calculations, FNPW ~ C are found as follows.
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

~ B ¼ $ð230; 099:088; 12; 366:329; 73; 730:964Þ

FNPW

~ C ¼ $ð220; 656:421; 26; 284:090; 80; 414:242Þ:

FNPW

Using Liou and Wang’s ranking method, for a moderately optimistic decision-maker,
w ¼ 0:5; we obtain
~ ¼ 2124:597
E0:5 ðAÞ ~ ¼ 17; 091:134
E0:5 ðBÞ ~ ¼ 28; 081:500:
E0:5 ðCÞ

Interpreting these results, we conclude that FNPW ~ C is strongly more important than
~ ~
FNPW B ; FNPW B is very strongly more important than FNPW ~ A ; and FNPW
~ C is absolutely
more important than FNPW A :~
Finally, the comparison matrix with respect to the software economic is obtained as given
in Table VI.
By applying the fuzzy AHP methodology, the weight vector for the software economic is
calculated as Wsoftware economic ¼ (0.00, 0.27, 0.73)T.

4.2 Calculations Using the Quality Model

As already mentioned, we use the Delphi technique to identify the importance weights of
software quality criteria and attributes. We can obtain the consensus rating for all
eight evaluators and transfer them to a spreadsheet as shown in Table VII(a) –(d). It is
important to note that the importance weights of criteria highly depend on the considered
sector and on the type of software. In our case, we required the evaluators to assess on various
client – server-based management software for SMEs.
By using the fuzzy values in Table IV, the fuzzy evaluation matrix relevant to the software
quality factor is given in Table VIII.
By applying the fuzzy AHP methodology described in the “Extent analysis method
on fuzzy AHP” section, the weight vector from Table VIII is calculated as
Wquality factors ¼ (0.21, 0.45, 0.34)T. We compute for all related criteria and we obtained the
final criteria weights given in Table IX.
At the second level of the evaluation procedure, the decision committee compares software
development strategies A, B and C under each of the quality attributes separately. These
results in the matrices are shown in Table X(a) –(l).
The combination of weights for strategic criteria, attributes and alternatives to determine
priority weights for strategic factor “software quality” is given in Table XI.

4.3 Evaluations of Strategic Factors

As we mentioned before, the economic and quality factors are both of the same importance
for the company. Based on the results obtained by the evaluation of strategic factors as shown
in Table XII, software development strategy C is selected as the most appropriate.
 Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

274
TABLE VII Relative importance of the software quality factors, criteria and attributes

(a) Relative importance of software quality factors

Design quality Performance quality Adaptation quality
Design quality
Performance quality Weakly more important Weakly more important
Adaptation quality Weakly more important

(b) In software design quality criteria, the relative importance of attributes

Correctness Maintainability Verifiability
Correctness Weakly more important Strongly more important
Maintainability Weakly more important
Verifiability

(c) In software performance quality criteria, the relative importance of attributes

Efficiency Integrity Reliability Usability Testability
Efficiency Equally important Weakly more important Weakly more important Weakly more important
Integrity Weakly more important Weakly more important Weakly more important
G. BÜYÜKÖZKAN et al.

Reliability Equally important Equally important

Usability Equally important
Testability

(d) In software adaptation quality criteria, the relative importance of attributes

Expandability Flexibility Portability Reusability
Expandability Very strongly more important
Flexibility Weakly more important Equally important Strongly more important
Portability Weakly more important Strongly more important
Reusability
 A FUZZY MULTI-CRITERIA DECISION 275

TABLE VIII The fuzzy evaluation matrix with respect to the software quality

Design quality Performance quality Adaptation quality

Design quality (1, 1, 1) (1/2, 2/3, 1) (1/2, 2/3, 1)
Performance quality (1, 3/2, 2) (1, 1, 1) (1, 3/2, 2)
Adaptation quality (1, 3/2, 2) (1/2, 2/3, 1) (1, 1, 1)

TABLE IX Composite priority weight for strategic software quality factor

Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

Criteria Local weights Attributes Local weights

Software design quality 0.21 Correctness 0.56
Maintainability 0.34
Verifiability 0.10

Software performance quality 0.45 Efficiency 0.24

Integrity 0.24
Reliability 0.17
Usability 0.17
Testability 0.18

Software adaptation quality 0.34 Expandability 0.29

Flexibility 0.34
Portability 0.34
Reusability 0.03

TABLE X Evaluation of software development strategies with respect to the software quality attributes

A B C
(a) Correctness
A (1, 1, 1) (2/5, 1/2, 2/3) (2/5, 1/2, 2/3)
B (3/2, 2, 5/2) (1, 1, 1) (1/2, 2/3, 1)
C (3/2, 2, 5/2) (1, 3/2, 2) (1, 1, 1)
The weight vector is calculated as WCorrectness ¼ (0.00, 0.43, 0.57)T

(b) Maintainability
A (1, 1, 1) (2/5, 1/2, 2/3) (1/2, 2/3, 1)
B (3/2, 2, 5/2) (1, 1, 1) (1, 3/2, 2)
C (1, 3/2, 2) (1/2, 2/3, 1) (1, 1, 1)
The weight vector is calculated as W Maintainability ¼ ð0:10; 0:56; 0:34ÞT

(c) Verifiability
A (1, 1, 1) (1/2, 2/3, 1) (2/5, 1/2, 2/3)
B (1, 3/2, 2) (1, 1, 1) (1/2, 1, 3/2)
C (3/2, 2, 5/2) (2/3, 1, 2) (1, 1, 1)
The weight vector is calculated as W Verifiability ¼ ð0:6; 0:39; 0:45ÞT

(d) Efficiency
A (1, 1, 1) (2/5, 1/2, 2/3) (1/2, 2/3, 1)
B (3/2, 2, 5/2) (1, 1, 1) (1, 3/2, 2)
C (1, 3/2, 2) (1/2, 2/3, 1) (1, 1, 1)
The weight vector is calculated as W Efficiency ¼ ð0:10; 0:56; 0:34ÞT

(e) Integrity
A (1, 1, 1) (1/3, 2/5, 1/2) (2/5, 1/2, 2/3)
B (2, 5/2, 3) (1, 1, 1) (1, 3/2, 2)
C (3/2, 2, 5/2) (1/2, 2/3, 1) (1, 1, 1)
The weight vector is calculated as W Integrity ¼ ð0:00; 0:62; 0:38ÞT
 276 G. BÜYÜKÖZKAN et al.

TABLE X – continued

A B C

(f) Reliability
A (1, 1, 1) (1/2, 2/3, 1) (1/2, 2/3, 1)
B (1, 3/2, 2) (1, 1, 1) (1, 3/2, 2)
C (1, 3/2, 2) (1/2, 2/3, 1) (1, 1, 1)
The weight vector is calculated as W Reliability ¼ ð0:21; 0:45; 0:34ÞT
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

(g) Usability
A (1, 1, 1) (1/3, 2/5, 1/2) (3/2, 2, 5/2)
B (2, 5/2, 3) (1, 1, 1) (1, 3/2, 2)
C (3/2, 2, 5/2) (1/2, 2/3, 1) (1, 1, 1)
The weight vector is calculated as W Usability ¼ ð0:23; 0:47; 0:3ÞT

(h) Testability
A (1, 1, 1) (1/2, 2/3, 1) (1/2, 2/3, 1)
B (1, 3/2, 2) (1, 1, 1) (1/2, 2/3, 1)
C (1, 3/2, 2) (1, 3/2, 2) (1, 1, 1)
The weight vector is calculated as W Testability ¼ ð0:21; 0:34; 0:45ÞT

(i) Expandability
A (1, 1, 1) (1/2, 2/3, 1) (2/5, 1/2, 2/3)
B (1, 3/2, 2) (1, 1, 1) (1/2, 2/3, 1)
C (3/2, 2, 5/2) (1, 3/2, 2) (1, 1, 1)
The weight vector is calculated as W Expandability ¼ ð0:10; 0:34; 0:56ÞT

(j) Flexibility
A (1, 1, 1) (2/5, 1/2, 2/3) (1/2, 2/3, 1)
B (3/2, 2, 5/2) (1, 1, 1) (1, 3/2, 2)
C (1, 3/2, 2) (1/2, 2/3, 1) (1, 1, 1)
The weight vector is calculated as W Flexibility ¼ ð0:10; 0:56; 0:34ÞT :

(k) Portability
A (1, 1, 1) (2/5, 1/2, 2/3) (2/5, 1/2, 2/3)
B (3/2, 2, 5/2) (1, 1, 1) (2/3, 1, 2)
C (3/2, 2, 5/2) (1/2, 1, 3/2) (1, 1, 1)
The weight vector is calculated as W Portability ¼ ð0:09; 0:46; 0:46ÞT :

(l) Reusability
A (1, 1, 1) (2/5, 1/2, 2/3) (1/2, 2/3, 1)
B (3/2, 2, 5/2) (1, 1, 1) (1, 3/2, 2)
C (1, 3/2, 2) (1/2, 2/3, 1) (1, 1, 1)
The weight vector is calculated as W Reusability ¼ ð0:10; 0:56; 0:34ÞT

5. CONCLUSIONS

Due to the difficulty that a decision-maker faces in the uncertainty and vagueness from
subjective perception and experience in the decision process, this study proposed to use a
selection framework based on fuzzy multi-criteria evaluation methodology to make suitable
decisions under a fuzzy environment. Both economic and quality factors were evaluated to
obtain the preference degree associated with each software development project strategy
alternative for selecting the most appropriate one. By using the proposed approach, the
ambiguities involved in the assessment data could be effectively represented and processed to
assure a more convincing and effective decision-making.
 Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

TABLE XI Summary combination of priority weights for “software quality” factor

Attributes of design
Correctness Maintainability Verifiability Alternative priority weight
Weight 0.56 0.34 0.10
A 0.00 0.10 0.16 0.05
B 0.43 0.56 0.39 0.47
C 0.57 0.34 0.45 0.48

Attributes of performance
Efficiency Integrity Reliability Usability Testability Alternative priority weight
Weight 0.24 0.24 0.17 0.17 0.18
A 0.10 0.00 0.21 0.23 0.21 0.14
B 0.56 0.62 0.45 0.47 0.34 0.50
C 0.34 0.38 0.34 0.30 0.45 0.36

Attributes of adaptation
Expandability Flexibility Portability Reusability Alternative priority weight
Weight 0.29 0.34 0.34 0.03
A 0.10 0.10 0.09 0.10 0.10
B 0.34 0.56 0.46 0.56 0.46
C 0.56 0.34 0.46 0.34 0.44
A FUZZY MULTI-CRITERIA DECISION

Strategic criteria of software quality

Design Performance Adaptation Alternative priority weight
Weight 0.21 0.45 0.34
A 0.05 0.14 0.10 0.11
B 0.47 0.50 0.46 0.48
C 0.48 0.36 0.44 0.41
277
 278 G. BÜYÜKÖZKAN et al.

TABLE XII Final evaluation of software development project strategy selection

Economic Quality Alternative priority weight

Weight 0.50 0.50
A 0.00 0.11 0.055
B 0.27 0.48 0.375
C 0.73 0.41 0.570
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

The proposed algorithm is simple but time-consuming. In the near future, we aim
to further develop an extended tool in which by conducting fuzzy assessments, the decision-
makers can obtain the final ranking of development strategies automatically.
In our second perspective, we want to apply the other multi-attribute evaluation methods such
as ELECTRE, DEA and TOPSIS (Triantaphyllou, 2000). These methods have been recently
developed to use in a fuzzy environment (Chen, 2000; Guo and Tanaka 2001; Leyva-López and
Fernández-González, 2003). Further research may include the application of these methods to
the software development strategy selection problem and the comparison of the results.

References
Ashrafi, N. (2002) “The impact of software process improvement on quality in theory and practice”, Information &
Management 2024, 1–14.
Boender, C.G.E., de Grann, J.G. and Lootsma, F.A. (1989) “Multicriteria decision analysis with fuzzy pairwise
comparison”, Fuzzy Sets and Systems 29, 133–143.
Bozdağ, C.H., Kahraman, C. and Ruan, D. (2003) “Fuzzy group decision making for selection among computer
integrated manufacturing systems”, Computers in Industry 51(1), 13– 29.
Buckley, J.J. (1985) “Fuzzy hierarchical analysis”, Fuzzy Sets and Systems 17, 233–247.
Buckley, J.J. (1987) “The fuzzy mathematics of finance”, Fuzzy Sets and Systems 21, 257– 273.
Chang, D-Y. (1992) “Extent analysis and synthetic decision”, Optimization Techniques and Applications 1, 352.
Chang, D-Y. (1996) “Applications of the extent analysis method on fuzzy AHP”, European Journal of Operational
Research 95, 649–655.
Chen, C.T. (2000) “Extensions of the TOPSIS for group decision-making under fuzzy environment”, Fuzzy Sets and
Systems 114, 1–9.
Cheng, C-H. (1996) “Evaluating naval tactical missile systems by fuzzy AHP based on the grade value of
membership function”, European Journal of Operational Research 96(2), 343 –350.
Cheng, C-H., Yang, K-L. and Hwang, C-L. (1999) “Evaluating attack helicopters by AHP based on linguistic
variable weight”, European Journal of Operational Research 116(2), 423–443.
Chiu, C.Y. and Park, C.S. (1994) “Fuzzy cash flow analysis using present worth criterion”, The Engineering
Economist 39(2), 113– 138.
Delbecq, A.L., Van de Ven, A.H. and Gustafson, D.H. (1975) Group Techniques for Program Planning (Scott,
Foresman and Company, Glenview, IL).
Deng, H. (1999) “Multicriteria analysis with fuzzy pairwise comparison”, International Journal of Approximate
Reasoning 21(3), 215– 231.
Glass, R.L. (1998) Software Runaways (Prentice-Hall, New Jersey).
Guo, P. and Tanaka, H. (2001) “Fuzzy DEA: a perceptual evaluation method”, Fuzzy Sets and Systems 119, 149 –160.
Hareton, K. and Leung, N. (2001) “Quality metrics for intranet applications”, Information & Management 38(3),
137– 152.
Hines, M.L. and Goerner, A.A. (1995) “Software quality: attributes and modalities”, Software Quality Management
III Vol. 2. Measuring and Maintaining Quality (Computational Mechanics Publications, Southampton),
pp 137–146.
Hoffman, T. (1999) “85% of IT departments fail to meet business needs”, Computer World 33(41), 24.
Jørgensen, M. (1999) “Software quality measurement”, Advances in Engineering Software 30(12), 907–912.
Kahraman, C. (2001) “Capital budgeting techniques using discounted fuzzy cash flows”, In: Ruan, D., Kacprzyk, J.
and Fedrizzi, M., eds, Soft Computing for Risk Evaluation and Management: Applications in Technology,
Environment and Finance (Physica Verlag, Heidelberg), pp 375–396.
Kahraman, C., Tolga, E. and Ulukan, Z. (2000) “Justification of manufacturing technologies using fuzzy benefit/cost
ratio analysis”, International Journal of Production Economics 66(1), 45–52.
Kahraman, C., Ruan, D. and Tolga, E. (2002) “Capital budgeting techniques using discounted fuzzy versus
probabilistic cash flows”, Information Sciences 42(1–4), 57–76.
Kenis, D., (1995) “Improving group decisions: designing and testing techniques for group decision support systems
applying Delphi principles”, Ph.D. Thesis, Utrech University.
 A FUZZY MULTI-CRITERIA DECISION 279

Kuo, R.J., Chi, S.C. and Kao, S.S. (2002) “A decision support system for selecting convenience store location
through integration of fuzzy AHP and artificial neural network”, Computers in Industry 47(2), 199–214.
Kwong, C.K. and Bai, H. (2003) “Determining the importance weights for the customer requirements in QFD using
a fuzzy AHP with an extent analysis approach”, IIE Transactions 35, 619–626.
Lee, M., Pham, H. and Zhang, X. (1999) “A methodology for priority setting with application to software
development process”, European Journal of Operational Research 118(2), 375– 389.
Leung, L.C. and Cao, D. (2000) “On consistency and ranking of alternatives in fuzzy AHP”, European Journal of
Operational Research 124(1), 102 –113.
Leyva-López, J.C. and Fernández-González, E. (2003) “A new method for group decision support based on
ELECTRE III methodology”, European Journal of Operational Research 148, 14–27.
Liou, T-S. and Wang, M-J. (1992) “Ranking fuzzy numbers with integral value”, Fuzzy Sets and Systems 50, 247 –255.
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

McConnell, S. (1996) Rapid Development (Microsoft Press, Redmond).

Paulk, C. (1995) The Capability Maturity Model: Guidelines for Improving the Software Process (Addison-Wesley,
Reading, MA).
Saaty, T.L. (1980) The Analytic Hierarchy Process (McGraw Hill Company, New York).
Schenk, K.D., Vitalari, N.P. and Davis, S.K. (1998) “Differences between novice and expert systems analysts: what
do we know and what do we do?”, Journal of Management Information Systems 15(1), 9– 51.
Stam, A., Minghe, S. and Haines, M. (1996) “Artificial neural network representations for hierarchical preference
structures”, Computers & Operations Research 23(12), 1191–1201.
Triantaphyllou, E. (2000) Multi-Criteria Decision Making Methods: A Comparative Study (Kluwer Academic
Publishers, London).
Van Laarhoven, P.J.M. and Pedrycz, W. (1983) “A fuzzy extension of Saaty’s priority theory”, Fuzzy Sets and
Systems 11, 229–241.
Verner, J.M., Overmyer, S.P. and McCain, K.W. (1999) “In the 25 years since the mythical man-month what we
learned about project management?”, Information and Software Technology 41(14), 1021–1026.
Wang, J. and Lin, Y.-I. (2003) “A fuzzy multicriteria group decision making approach to select configuration items
for software development”, Fuzzy Sets and Systems 134(3), 343–363.
Ward, T.L. (1989) “Fuzzy discounted cash flow analysis”, In: Evans, G.W., Karwowski, W. and Wilhelm, M.R. (Eds.)
Applications of Fuzzy Set Methodologies in Industrial Engineering 2, 91– 112.
Weck, M., Klocke, F., Schell, H. and Rüenauver, E. (1997) “Evaluating alternative production cycles using the
extended fuzzy AHP method”, European Journal of Operational Research 100(2), 351 –366.
Yu, C.-S. (2002) “A GP-AHP method for solving group decision-making fuzzy AHP problems”, Computers &
Operations Research 29(14), 1969–2001.
Zadeh, L.A. (1965) “Fuzzy sets”, Information & Control 8, 338– 353.
Zeist, R.H.J. and Hendriks, P.R.H. (1996) “Specifying software quality with the extended ISO model”, Software
Quality Management IV — Improving Quality (Mechanical Engineering Publishing, London), pp 145–160.
Zhu, K-J., Jing, Y. and Chang, D-Y. (1999) “A discussion on extent analysis method and applications of fuzzy AHP”,
European Journal of Operational Research 116(2), 450–456.

Gülçin Büyüközkan, after completing her undergraduate studies in

Industrial Engineering at Istanbul Technical University obtained her
M.Sc. degree in Industrial Engineering from ENSGI/INPG in France
(1996) and also from Bosphorus University in Turkey (1997).
In November 1999, she completed her Ph.D. studies in Industrial
Engineering in INPG, France, and since then has worked as an Assistant
Professor in the Industrial Engineering Department at Galatasaray
University, Turkey. Her post-doctoral studies mainly focused on effective
decision-making for organizational performance improvement and their
industrial applications. Miss Büyüközkan has published several
international conference and journal papers.

Cengiz Kahraman is an Associate Professor in the Department of

Industrial Engineering, Istanbul Technical University (ITU), Turkey.
He received his B.Sc., M.Sc. and Ph.D. degrees in Industrial Engineering
from ITU. He is currently the Assistant Head of the Department and his
research areas include engineering economics, statistics, quality control
and the applications of fuzzy sets theory on these areas. He has published
several international conference papers, journal papers, and book
chapters.
 280 G. BÜYÜKÖZKAN et al.

Da Ruan (BS, Applied Mathematics, Fudan University, Shanghai, 1983;

Ph.D., Mathematics Gent University, Belgium, 1990) is senior researcher
with the Belgian Nuclear Research Centre (SCKCEN) working in the
areas of applied computational intelligence for uncertainty analysis,
information/sensor fusion, decision support systems, robotics, nuclear
power plants and safety-related engineering fields.
Downloaded by [Istanbul Technical University] at 07:47 14 November 2014

View publication stats