Available online at www.sciencedirect.

com

Expert Systems
with Applications
Expert Systems with Applications 36 (2009) 3346–3354
www.elsevier.com/locate/eswa

Evaluating knowledge management capability of organizations:

a fuzzy linguistic method
Zhi-Ping Fan a, Bo Feng a,*, Yong-Hong Sun b, Wei Ou a
a
Department of Management Science and Engineering, School of Business Administration, Northeastern University, Shenyang 110004, China
b
Department of Information Systems, City University of Hong Kong, Kowloon, Hong Kong, China

Abstract

Knowledge management capability (KMC) is the source for organizations to gain the sustainable competitive advantage. KMC eval-
uation is a required work with strategic significance. However it still has not been addressed in the existing literatures. So the objective of
this study is to investigate a fuzzy multiple attributes decision-making method (FMADM) for evaluating KMC. In this paper, a frame-
work for evaluating KMC is presented, which includes two parts, one is an evaluation hierarchy with attributes, the other a judgment
matrix model with two dimensions to identify the evaluation results of KMC. Then, a fuzzy linguistic approach is proposed to evaluate
the KMC of organizations. The evaluation results of KMC obtained through the proposed approach are objective and unbiased due to
two reasons. Firstly, the results are generated by a group of experts in the presence of motile attributes. Secondly, the fuzzy linguistic
approach employed in this paper has more advantage to reduce distortion and losing of information than other fuzzy linguistic
approaches. Through evaluation result of KMC, managers could judge the necessity to improve the KMC and determine which dimen-
sion of KMC is the most needed direction to improve. Additionally, an example is used to illustrate the availability of the proposed
method.
Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Knowledge management (KM); Knowledge management capability (KMC); Linguistic assessment information; Judgment matrix model;
2-Tuple; Fuzzy linguistic approach

1. Introduction nately, many KM projects are, in reality, information man-

Knowledge management (KM) has been described for
its possible role in creating sustained competitive advanta-
ges for organizations (Chuang, 2004; Grant, 1996; Johann-
essen & Olsen, 2003; Nonaka & Takeuchi, 1995). The
contributions of KM to competitive advantage may
include: improved ability of innovation, improved coordi-
nation of efforts and rapid commercialization of new prod-
ucts. Other contributions may include: the ability to
anticipate surprise, responsiveness to market change, and
reduced redundancy of information/knowledge. So, many
organizations are making extensive KM efforts. Unfortu-
*
Corresponding author. Tel.: +86 24 83687753; fax: +86 24 23891569.
E-mail addresses: zpfan@mail.neu.edu.cn (Z.-P. Fan), neu_fengbo@
163.com, msbofeng@yahoo.cn (B. Feng).
agement ones. When these projects yield some
consolidation of data but little innovation in products
and services, the concept of KM is cast in doubt (Gold,
Malhotra, & Segars, 2001). The main reason for this prob-
lem is that organizations may not identify and assess the
preconditions that the efforts to KM are necessary. There-
fore, organizations cannot understand the success and fail-
ure of KM within organizations. These preconditions are
described broadly as ‘capability’ or ‘resources’ within the
organizational behavior literature (Kelly & Amburgey,
1991; Law, Wong, & Mobley, 1998; Leonard, 1995).
There has been much research dealing with KM capabil-
ity (KMC). Desouza (2003) argued that the ideal organiza-
tion with well-matured KMC can ensure the identification,
distribution, protection, application and destruction of
knowledge. Therefore, KMC is the key to preempting an

0957-4174/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eswa.2008.01.052
 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354 3347

organizational crisis. Lubit (2001) argued that tacit knowl- hierarchy, a judgment matrix model with two dimensions
edge and superior KMC are now the keys to sustainable is constructed, one dimension is infrastructure capability
competitive advantage in many industries. Liu, Wen, and and the other is process capability. The evaluation result
Tsai (2004) examined the association between KMC and of KMC can be visualized in the matrix model, which is
competitiveness by empirical study. The result reveals that convenient for managers to identify the KMC of
KMC has a tremendous effect on organizational competi- organizations.
tiveness. KMC is considered more than a catch-all for
information and knowledge. It is a tool for maintaining
2.1. An evaluation hierarchy for KMC
information and knowledge that will help employees to
work more efficiently (Liu et al., 2004). Collinson (2001)
The contents of KMC presented by Gold et al. (2001)
emphasized the significance of contextual factors for trans-
and Chuang (2004) were expressed with two dimensions
ferring some KM practices by case study. Bresnen, Edel-
and seven attributes. One dimension is infrastructure capa-
man, Newell, Scarbrough, and Swan (2003) examined the
bility, including attributes of technology, structure and cul-
significance of social factors in enhancing KMC in project
ture. The other dimension is process capability, including
environments by case study. Gold et al. (2001) and Chuang
attributes of acquisition, conversion, application and secu-
(2004) presented and validated the framework for analysis
rity. Gold et al. (2001) argued that knowledge capabilities
of KMC using different attributes. Thus, many efforts have
were additive in nature according to the empirical research.
been made to emphasize the significance of KMC, and ana-
Infrastructure capability was a sum of technological, struc-
lyze and explore the attributes of KMC. However, the eval-
tural and cultural capability. Likewise, process capability
uation of KMC with the qualitative multi-attributes has
was an additive effect of acquisition, conversion, applica-
seldom been addressed.
tion and protection capability. Thus, KMC can be addi-
Indeed, there are many approaches that can be used to
tively determined by its dimensions and extending
evaluate the KMC. For example, scoring tool may be the
attributes. For convenience, the infrastructure capability
simplest approach to evaluate the KMC. However, usually,
is represented as X. Its attributes are represented as X1,
most experts can not give exact numerical values to express
X2, and X3 accordingly. Likewise, the process capability
their opinions based on human perception. More realistic
is represented as Y. Its attributes are represented as Y1,
measurement is to use linguistic assessments instead of
Y2, Y3, and Y4. The evaluation hierarchy of KMC is shown
numerical values (Beach, Muhlemann, Price, Paterson, & as Fig. 1. The grades of importance of these attributes
Sharp, 2000; Gerwin, 1993; Herrera & Herrera-Viedma, depend on the industry to which an organization belongs
2000; Kacprzyk, 1986; Vokurka & O’Leary-Kelly, 2000). and the strategy that the organization implements. Fur-
Attributes can be measured as linguistic labels (or terms) thermore, in order to facilitate experts to provide precise
such as ‘very high’, ‘high’, ‘middle’, ‘low’, and ‘very low’ judgments, the details on attributes of each dimension of
(Wang & Chuu, 2004). After Zadeh (1965) introduced KMC are also described in Table 1.
fuzzy set theory to deal with vague problems, linguistic
labels have been used within the framework of fuzzy set
theory (Zadeh, 1975a, 1975b, 1976) to handle the ambigu- 2.2. A judgment matrix model for KMC
ity in evaluation data and the vagueness of linguistic
expression (Wang & Chuu, 2004). From the above, it is easy to see that the KMC of orga-
Therefore, the purpose of this study is to establish an nizations is related to two dimensions, infrastructure capa-
evaluation framework of KMC for organizations and to bility and process capability. Both of them are very
investigate a fuzzy linguistic approach to evaluate the important and any one can not be missing. More specific,
KMC in a fuzzy environment. Section 2 presents an evalu-
ation framework of KMC for organizations, in which, the
dimensions and attributes of KMC are introduced and a
judgment matrix model is presented. Based on the charac-
teristics of dimensions and attributes discussed in Section 2,
a fuzzy linguistic approach is then proposed to evaluate the
KMC of organizations in Section 3. Section 4 illustrates the
proposed method with an example.

2. An evaluation framework for KMC

In this section, we will present an evaluation framework

for KMC. The framework consists of two parts, an evalu-
ation hierarchy and a judgment matrix model for KMC. In
the hierarchy, the attributes for evaluating KMC are final- Fig. 1. The evaluation hierarchy of KMC (Chuang, 2004; Gold et al.,
ized through literature review. Based on the evaluation 2001).
3348 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354
Table 1
The attributes of infrastructure capability and process capability (Chuang, 2004; Gold et al., 2001)
Dimensions Attributes
Infrastructure capability Technology
Structure
Culture
Process capability Acquisition
Conversion
Application
Security
if one organization spends a vast investment on IT and
takes a hard endeavor on culture construction, but repre-
sents badly in knowledge acquisition, conversion, applica-
tion and security, namely, it has strong infrastructure
capability and week process capability, then the utilization
efficiency of IT equipments may be not high and the KMC
of organizations may be not strong. On the other hand, if
capabilities of knowledge acquisition, conversion, applica-
tion and security is strong in an organization, but without
the supporting of information technologies and culture,
then the exertion of process capability will be restricted.
Therefore, the KMC of organizations depends on the two
dimensions, infrastructure capability and process capabil-
ity, i.e., both of them must be balanced and improved. If
we only consider the collective evaluation result of KMC,
then the information of one dimension may be submerged
by that of the other.
Furthermore, KMC of an organization can be identified
by a judgment matrix model, which is graphically shown as
Fig. 2. In Fig. 2, horizontal and vertical axes denote dimen-
sions of process capability and infrastructure capability,
respectively. So the KMC of an organization can be shown
by one point of Cartesian coordinate system in Fig. 2. The
point ‘Middle’ represents the middle degree, and ‘Definitely
High’ and ‘Definitely Low’, respectively represents the
highest degree and the lowest degree. In the matrix model,
the points in quadrant I represent both of the capabilities
Fig. 2. The judgment matrix model for KMC.
Meanings
Integrate previously fragmented flows of information/knowledge in organization
Important in leveraging technological architecture
Employee interaction should be encouraged, both formally and informally
Acquisition-oriented KM processes are those oriented toward obtaining knowledge
Make existing knowledge useful
Actual use of the knowledge
Protect the knowledge from inappropriate or illegal use or theft
are low, namely, KMC is low. The points in quadrants II
and III represent one capability is high while the other is
low, so the KMC need to be improved with respect to
the low dimension. Only the points in quadrant IV repre-
sents both of the capabilities are high, namely, KMC is
high. It can be seen that quadrant IV in Fig. 2 is the high
and effective region of KMC, which is marked by
diagonals.
Generally, the evaluation of KMC requires group opin-
ions from multiple experts. They are responsible for pro-
viding assessment information for the performance and
importance of each attributes. From Table 1, it is obvious
that all attributes of KMC are qualitative, so the easiest
way for experts to express their opinions is to use linguistic
variables (namely, linguistic terms or labels) such as ‘‘Very
High”, ‘‘High” or ‘‘Middle”. Therefore, the situation that
experts express their opinions by use of linguistic assess-
ment information is considered in this study. Because the
preference information delivered by linguistic variables is
fuzzy, a fuzzy linguistic approach is then proposed to eval-
uate the KMC of organizations.
3. Fuzzy linguistic approach
3.1. Linguistic assessments
In the fuzzy linguistic approaches, linguistic variables
are used to denote words or sentences of a natural language
(Zadeh, 1975a, 1975b, 1976). The approaches are appropri-
ate for many evaluation problems in which information
may be qualitative, or quantitative information may not
be stated precisely, since either it is unavailable or the cost
of its determination is excessive, such that an ‘approximate
value’ suffices (Herrera & Herrera-Viedma, 2000; Wang &
Chuu, 2004).
Usually, most experts tend toward providing linguistic
assessments rather than exact numerical values to express
their opinions for evaluating the KMC. They will give
the evaluations in the form of linguistic terms in the pres-
ence of qualitative attributes. The information should
include the performance and importance of each attribute.
The importance of each attribute should be given accord-
ing to the KM strategy of organizations. Then the infra-
structure and process capability can be measured with the
performance and importance of each attribute.
 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354
As mentioned above, the rating of performance and
grade of importance should be measured for each attribute.
Therefore, both were scored on a nine-rank scale, as shown
in Table 2. Let S = {S0, S1, . . ., S8} be a finite and totally
ordered term set with odd cardinalities, where the middle
label, S4, represents ‘average’, and the remaining terms
are placed symmetrically around S4. There exists the fol-
lowing properties (Herrera, Herrera-Viedma, & Verdegay,
1998).
1. The set is ordered: Si‘‘ P ”Sj if i P j, where ‘‘P”
denotes ‘greater than or equal to’.
2. There is a negation operator: Neg (Si) = Sj such that
j = 8i, where 8+1 is the cardinality of the set S.
3. Maximization operator: MAX (Si, Sj) = Si if Si‘‘ P ”Sj.
4. Minimization operator: MIN (Si, Sj) = Sj if Si‘‘ P ”Sj.
The nine linguistic labels in S = {S0, S1, . . ., S8} were
specified. This paper considers a situation in which experts
can perfectly distinguish among the set of labels under a
similar conception, and can use linguistic labels to express
their opinions (Wang & Chuu, 2004).
3.2. 2-Tuple linguistic representation model
The approaches for dealing with linguistic information
can be classified into three categories (Herrera & Martinez,
2000). The first one is based on the Extension Principle
(Adamopoulos & Pappis, 1996; Bordogna & Passi, 1993;
Herrera & Herrera-Viedma, 2000; Torra, 1997). It makes
operations on the fuzzy numbers that support the seman-
tics of the linguistic terms. The second one is the symbolic
approach (Bordogna, Fedrizzi, & Passi, 1997; Herrera, Her-
rera-Viedma, & Verdegay, 1996, 1998). It makes computa-
tions on the indexes of the linguistic terms. The third one is
based on 2-tuple fuzzy linguistic representation model
(Herrera & Martinez, 2000, 2001). In the former two
approaches, the results usually do not exactly match any
of the initial linguistic terms, and then an approximation
process must be developed to express the result in the initial
expression domain. This produces the consequent loss of
information and hence the lack of precision (Bordogna &
Passi, 1993). The third category of approach overcomes
the above weakness. The model represents the linguistic
Table 2
Linguistic labels for rating of performance and grade of importance
Nine ranks of rating of performance Nine ranks of grade of importance
S0 = DL: definitely low S0 = DL: definitely low
S1 = VL: very low S1 = VL: very low
S2 = L: low S2 = L: low
S3 = ML: more or less low S3 = ML: more or less low
S4 = M: middle S4 = M: middle
S5 = MH: more or less high S5 = MH: more or less high
S6 = H: high S6 = H: high
S7 = VH: very high S7 = VH: very high
S8 = DH: definitely high S8 = DH: definitely high
3349
assessment information with a pair of values, which is
called 2-tuple, composed by a linguistic term and a
number. The main advantage of this representation is to
be continuous in its domain, so it can express any counting
of information in the universe of the discourse. The
detailed comparative results of the three methods can be
obtained from the research of Herrera and Martinez
(2000). Through the comparison, it is easy to find out that
the third category of approach is more convenient and
precise when dealing with fuzzy linguistic assessment
information.
2-Tuple linguistic representation model, presented in lit-
eratures (Herrera & Martinez, 2000, 2001), is based on the
concept of symbolic translation. It is used for representing
the linguistic assessment information by means of 2-tuple
(Si, a), where Si is labels from predefined linguistic term
set S, a is the value of symbolic translation, and
a 2 [0.5, 0.5).
Let Si 2 S be a linguistic label. Then the function h used
to obtain the corresponding 2-tuple linguistic information
of Si is defined as
h : S ! S  ½0:5; 0:5Þ
ð1Þ
hðS i Þ ¼ ðS i ; 0Þ; S i 2 S
Let S = {S0, S1, . . ., ST} be a linguistic term set, b 2 [0, T]
is a number value representing the aggregation result of lin-
guistic symbolic. Then the function D used to obtain the 2-
tuple linguistic information equivalent to b is defined as
D : ½0; T Þ ! S  ½0:5; 0:5Þ

Si; i ¼ Round ðbÞ ð2Þ
DðbÞ ¼ ðS i ; aÞ; with
a ¼ b  i; a 2 ½0:5; 0:5Þ
where Round is the usual round operation. Si has the clos-
est index label to b and a is the value of the symbolic trans-
lation. If S is a linguistic term set, S = {S0, S1, . . ., ST},
(Si,a) is 2-tuple linguistic information, then there exists a
function D1, which is able to transform 2-tuple linguistic
information into its equivalent numerical value b 2 [0, T].
The function D1 is defined as
D1 : S  ½0:5; 0:5Þ ! ½0; T 
ð3Þ
D1 ðS i ; aÞ ¼ i þ a ¼ b
If (Si, a1) and (Sj, a2) are two linguistic 2-tuples, they should
have the following properties:
1. The set is ordered: if i P j, then (Si, a1)‘‘ > ”(Sj, a2),
where ‘‘>” denotes ‘greater than’ If i = j, then
(i) if a1 > a2, then (Si, a1)‘‘ > ”(Sj, a2);
(ii) if a1 = a2, then (Si,a1)‘‘ = ”(Sj,a2), where ‘‘=”
denotes ‘equal to’;
(iii) if a1 < a2, then (Si, a1)‘‘ < ”(Sj, a2), where ‘‘<”
denotes ‘less than’.
2. There exists a negation operator: Neg ((Si, a)) = D
(T  (D1(Si, a))), such that, where T + 1 is the cardinal-
ity of the set S.
3350 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354

3. Maximization operator: MAX {(Si, ai), (Sj, aj)} = (Si, ai) Step 1: Transform the linguistic terms into the form of 2-
if (Si, ai)‘‘ P ”(Sj, aj). tuple linguistic. By the transformation function h
4. Minimization operator: MIN {(Si, ai), (Sj, aj)} = (Sj, aj) defined above, wr, prm, qrn, urm, and vrn can be
if (Si, ai)‘‘ P ”(Sj, aj). transformed into (wr, 0), (prm, 0), (qrn, 0), (urm, 0)
and (vrn, 0), respectively.
Let (b1, a1), (b2, a2), . . ., (bm, am) be a group of linguistic 2- Step 2: By the operator Be in Eq. (4), the average grades of
tuples to be aggregated, then 2-tuple arithmetic mean oper- importance of attributes Xm and Yn can be
ator Be is defined as obtained, where m = 1, 2, 3 and n = 1, 2, 3, 4. The
!
X m
1 aggregated grades of importance of attributes that
B ¼ ð
e
aÞ ¼ D
b;  D ðbi ; ai Þ ; 
1
b 2 S; 
a 2 ½0:5; 0:5Þ are obtained from the l experts’ linguistic assess-
i¼1
m
ment information can be expressed as follows:
ð4Þ !
Xl
1 1
Let (b1, a1), (b2, a2), . . ., (bm, am) be a group of linguistic 2- ðpm ; am Þ ¼ D D ðprm ; arm Þ ;
l
tuples to be aggregated, R ¼ ððr1 ; a01 Þ; ðr2 ; a02 Þ; . . . ; ðrm ; a0m ÞÞT r¼1

be its corresponding 2-tuple weighted vector, then 2-tuple m ¼ 1; 2; 3; pm 2 S; am 2 ½0:5; 0:5Þ ð6Þ
^ e is defined as !
weighted average operator B ! Xl
1 1
Pm
½D 1
ðr i ; a 0
Þ  D1
ðb i ; a i Þ ðqn ; an Þ ¼ D D ðqrn ; arn Þ ;
B e ^
^ ¼ ðb; ^ aÞ ¼ D i¼1
Pm 1 i
; r¼1
l
0
i¼1 D ðri ; ai Þ
n ¼ 1; 2; 3; 4; qn 2 S; an 2 ½0:5; 0:5Þ ð7Þ
^
b 2 S; ^
a 2 ½0:5; 0:5Þ ð5Þ
where (pm, am) and (qn, an) denote the aggregated
3.3. An approach for evaluating the KMC grades of importance of attributes of dimensions
X and Y, respectively.
In this section, according to the evaluation hierarchy of Step 3: By operator B ^ e in Eq. (5), aggregate the expert
KMC shown in Fig. 1 and the details shown in Table 1, firstly, evaluation information of attributes into the
we give the description of evaluation problem of KMC, and group evaluation information. Concretely, respec-
then the approach for evaluating the KMC is proposed by tively aggregate the grade of importance of experts
using the above 2-tuple linguistic representation model. (wr, 0) with the rating of performance of attribute
The experts use the linguistic term from the set (urm, 0) and (vrn, 0) by
S = {S0, S1, . . ., S8} to express their preference on the grade Pl !
1 1
of importance and rating of performance with regard to r¼1 ½D ðwr ; ar Þ  D ðurm ; arm Þ
ðum ; am Þ ¼ D Pl 1
attributes. In order to describe this problem, we have the r¼1 D ðwr ; ar Þ
following notation definitions. m ¼ 1; 2; 3; um 2 S; am 2 ½0:5; 0:5Þ; ð8Þ
Pl !
1 1
Er the rth expert (r = 1, 2, . . ., l); r¼1 ½D ðwr ; ar Þ  D ðvrn ; arn Þ
ðvn ; an Þ ¼ D Pl 1
Xm the mth attribute in dimension X (m = 1, 2, 3, see r¼1 D ðwr ; ar Þ
Fig. 1); n ¼ 1; 2; 3; 4; vn 2 S; an 2 ½0:5; 0:5Þ ð9Þ
Yn the nth attribute in dimension Y (n = 1, 2, 3, 4, see
Fig. 1); ^ e in Eq. (5), obtain the overall per-
Step 4: By operator B
wr the grade of importance of expert Er
formances of dimensions X and Y. Concretely,
(r = 1, 2, . . ., l);
integrate the grade of importance of attribute Xm
prm the grade of importance of attributes Xm provided
in Eq. (6) and the rating of performance of attri-
by expert Er (r = 1, 2, . . ., l; m = 1, 2, 3);
bute Xm in Eq. (8) by
qrn the grade of importance of attributes Yn provided
by expert Er (r = 1, 2, . . ., l; n = 1, 2, 3, 4); P3 !
1 1
urm the rating of performance of attribute Xm provided m¼1 ½D ðp m ; am Þ  D ðum ; am Þ
by expert Er (r = 1, 2, . . ., l; m = 1, 2, 3); ðuIC ; aIC Þ ¼ D P3 1
;
m¼1 D ðp m ; am Þ
vrn the rating of performance of attribute Yn provided
by expert Er (r=1, 2, . . ., l; n = 1, 2, 3, 4). uIC 2 S; aIC 2 ½0:5; 0:5Þ ð10Þ

Integrate the grade of importance of attributes Ym in Eq. (7)

Therefore, the following section of this paper will intro- and the rating of performance of attribute Ym in Eq. (9) by
duce the approach for evaluating the KMC by multiple !
P4 1 1
experts. This approach is used to aggregate the information n¼1 D ½ðqn ; an Þ  D ðvn ; an Þ
of performance of infrastructure capability and process ðvPC ; aPC Þ ¼ D P4 1
;
n¼1 D ðqn ; an Þ
capability, respectively. The calculating steps of the
approach are presented as follows. vPC 2 S; aPC 2 ½0:5; 0:5Þ ð11Þ
 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354 3351

The 2-tuple linguistic labels (vIC, aIC) and (vPC, aPC) rep-
resent overall performances of infrastructure capability and
process capability, respectively. According to the evaluat-
ing results of two dimensions, whether an organization
should improve its KMC or not could be determined by
the judgment matrix model presented in Section 2.
4. Illustrative example
Baosight is one of the top five software companies in
China, CMM/CMMi Level 5 accredited and is awarded
the Most Competent IT Service Provider in China in
2006. It owns 3400 employees, about 25% have Mater or
Ph.D. degree, and 75% have Bachelor degree. More than
500 of them have studied or have been trained abroad
and some 1000 involved in overseas projects. Baosight pro-
vides its clients with all-round, whole life cycle software
development design service, spanning from design, devel-
opment, coding, testing, and maintenance to application
development. Its IT Software Service is featured mainly
in the areas of Embedded Development, Mainframe Devel-
opment, Business Intelligence (BI), Collaborative Com-
merce, Manufacturing Execution System (MES) and
Enterprise Resource Planning (ERP).
Baosight is a typical knowledge-intensive enterprise. Its
employees mostly are knowledge workers, and its soft-
ware products are knowledge products. KM is a key tool
to obtain sustainable competitive advantage. Baosight
implements KM strategy and develops collaborative KM
platform to support internal management and operation
as well as external service. KM in Baosight essentially
involves project knowledge management; customer knowl-
edge management and knowledge push service. Baosight
takes promoting KMC as core competence source to
survive furious market competitions. It conducts
assessment on KMC to know the KMC level and judge
which dimension and aspects of KMC should be
improved.
To implement the evaluation task of KMC, three con-
cerned groups of respondents, including employees, exter-
nal experts, and senior managers, are organized to
evaluate to the KMC of Baosight. There are three respon-
dents in each group. Firstly, the mission, objective and KM
strategy of Baosight are stated clearly. Secondly, the attri-
butes of KMC are explained to respondents in great
details, in order that the respondents could provide objec-
tive and precise responses. Then each respondent is
requested to fill in a questionnaire. Their evaluation opin-
ions on the KMC of Baosight can be transformed accord-
ing to nine-rank term set (see Table 2), and the related
results are shown in Tables 3 and 4. The proposed
approach is used to evaluate the KMC.
Through the transformation function h in Eq. (1), trans-
form the linguistic variables in Tables 3 and 4 into the 2-
tuple linguistic form. The transformed results are shown
in Tables 5 and 6.

Table 3
The evaluation information of grades of importance of attributes
Dimensions Attributes Respondents’ opinions
Employees External experts Senior managers
E1 E2 E3 E4 E5 E6 E7 E8 E9
Infrastructure capability X1 M MH M M MH M H H VH
X2 VH VH DH L M ML MH M M
X3 H MH MH ML ML VL M VH DH
Process capability Y1 MH DH MH H MH H VH H M
Y2 DH VH MH M MH M MH H H
Y3 H VH VH H H MH VH M M
Y4 DH MH MH MH H M MH H MH

Table 4
The evaluation information of ratings of performance of attributes
Dimensions Attributes Respondents’ opinions
Employees External experts Senior managers
E1 E2 E3 E4 E5 E6 E7 E8 E9

Infrastructure capability X1 DH MH MH MH H M MH H MH
X2 VL H DL L M ML VL M ML
X3 H MH MH ML ML VL ML ML VL
Process capability Y1 DH VH MH M MH M M MH M
Y2 VH VH DH L M ML MH M M
Y3 H MH H ML ML VL DH VH MH
Y4 M ML M VH VH DH ML ML VL
3352 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354

Table 5
The transformed evaluation information of grades of importance of attributes
Attributes Respondents
E1 E2 E3 E4 E5 E6 E7 E8 E9
X1 (H, 0) (MH, 0) (VH, 0) (ML, 0) (MH, 0) (H, 0) (H, 0) (H, 0) (VH, 0)
X2 (ML, 0) (VH, 0) (M, 0) (MH, 0) (ML, 4) (H, 0) (MH, 0) (M, 0) (M, 0)
X3 (DH, 0) (MH, 0) (MH, 0) (DH, 0) (NH, 0) (DH, 0) (DH, 0) (VH, 0) (DH, 0)
Y1 (MH, 0) (MH, 0) (DH, 0) (H, 0) (MH, 0) (H, 0) (M, 0) (H, 0) (M, 0)
Y2 (H, 0) (NH, 0) (MH, 0) (NH, 0) (MH, 0) (M, 0) (MH, 0) (H, 0) (H, 0)
Y3 (DH, 0) (DH, 0) (DH, 0) (NH, 0) (H, 0) (DH, 0) (NH, 0) (H, 0) (H, 0)
Y4 (M, 0) (MH, 0) (DH, 0) (DH, 0) (NH, 0) (DH, 0) (MH, 0) (H, 0) (DH, 0)

Table 6
The transformed evaluation information of ratings of performance of attributes
Attributes Respondents
E1 E2 E3 E4 E5 E6 E7 E8 E9
X1 (H, 0) (MH, 0) (M, 0) (VH, 0) (VH, 0) (VH, 0) (H, 0) (H, 0) (H, 0)
X2 (M, 0) (ML, 0) (VL, 0) (L, 0) (M, 0) (L, 0) (VL, 0) (ML, 0) (L, 0)
X3 (M, 0) (MH, 0) (VH, 0) (ML, 0) (MH, 0) (L, 0) (VH, 0) (L, 0) (H, 0)
Y1 (MH, 0) (ML, 0) (MH, 0) (H, 0) (MH, 0) (H, 0) (VH, 0) (H, 0) (ML, 0)
Y2 (DH, 0) (VH, 0) (MH, 0) (DH, 0) (MH, 0) (DH, 0) (MH, 0) (H, 0) (H, 0)
Y3 (H, 0) (DH, 0) (VH, 0) (H, 0) (H, 0) (MH, 0) (VH, 0) (DH, 0) (DH, 0)
Y4 (DH, 0) (MH, 0) (MH, 0) (MH, 0) (H, 0) (VH, 0) (MH, 0) (H, 0) (MH, 0)

Table 7
The aggregated results of the KMC of Baosight
KMC Mean grade of importance Mean performances
of attributes of attributes
X (H, 0.47) X1 (H, 0.33) X1 (H, 0)
X2 (M, 0.33) X2 (ML, 0.44)
X3 (VH, 0.11) X3 (MH, 0.44)
Y (MH, 0.07) Y1 (MH, 0.44) Y1 (MH, 0.11)
Y2 (H, 0.33) Y2 (H, 0.44)
Y3 (VH, 0.11) Y3 (VH, 0.12)
Y4 (VH, 0.44) Y4 (H, 0.12)

DH
Fig. 4. The evaluation results of KMC Baosight.
VH
MH
(MH, 0.07). The evaluation results of each attribute in
H
Table 7 can be vividly shown by Fig. 3. Simultaneously,
M
the evaluation results of two dimensions is visualized by
L the judgment matrix model as shown in Fig. 4.
ML In Fig. 4, the performance of KMC of Baosight is
VL located in the quadrant IV in the judgment matrix model,
DL so the KMC of Baosight is high. The performance of infra-
X1 X2 X3 Y1 Y2 Y3 Y4
structure capability is closed to H, while the performance
Fig. 3. The evaluation results of each attribute of KMC of Baosight. of process capability is closed to MH, i.e. the former is
lower than the latter. Based on further analysis in Fig. 4,
we can find out that the performance of structure and cul-
Suppose that the opinions of respondents have the equal ture are relatively lower than that of other attributes. The
importance. By Eqs. (6)–(8) the aggregated results are KMC of Baosight is clear to the top managers. Then top
obtained and they are presented in Table 7. managers could know the priority of aspects needed to be
In Table 7, the performance of infrastructure capability improved. Therefore, they can take corresponding actions
is (H, 0.47) and performance of process capability is to enhance the KMC effectively and efficiently.
 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354
5. Conclusions
The proposed fuzzy linguistic method based on 2-tuple
linguistic representation model has the advantages that
include avoiding loss and distortion of experts’ assessment
information, obtaining the computation results as linguistic
labels and simplifying the calculation process. It is appro-
priate for the situations in which assessment information
is qualitative, or the precise quantitative information is
unavailable or the cost of its computation is too high.
The approach seems to be complex, but the calculation
process and principle are actually very easy. The compara-
tive analyses between this fuzzy linguistic approach and
others are illustrated in detail in the research of Herrera
and Martinez (2000). However, the approach is limited in
that experts must perfectly distinguish the set of labels
under a similar conception, and must use linguistic labels
to express their opinions.
The above method with the group evaluation structure
in the presence of multiple attributes, used to evaluate
the KMC of organizations, is very useful in KM initiation.
If the KMC is too low according to the evaluation results,
it should be improved until acceptable. The dimensions of
KMC needed improvements can be determined by the
judgment matrix model.
Some distinguished contributions of this study are as
follows:
1. Based on the evaluation hierarchy, a judgment matrix
model for KMC is presented. The evaluation results of
the KMC can be visualized in the matrix model. Thus,
the manager could easily judge the result by it. Only
when the result locates at the quadrant IV, the KMC
of an organization is high. According to the evaluation
result, managers could know the necessity to promote
the KMC and determine which dimension should be
improved.
2. A multiple criteria fuzzy linguistic approach with 2-tuple
linguistic representation model is proposed to evaluate
the KMC of organizations. The 2-tuple linguistic repre-
sentation model is newly developed and has been
applied successfully to several areas such as engineering
evaluation, SCM performance evaluation and compe-
tences evaluation, etc.
Acknowledgements
This work was partly supported by the National Science
Fund for Distinguished Young Scholars of China (Project
No. 70525002), National Science Fund for Excellent Inno-
vation Research Group of China (Project No. 70721001)
and the Key Laboratory of Integrated Automation of Pro-
cess Industry (Northeastern University), Ministry of Edu-
cation, China (Project No. JCLL-01-05). Gratitude is
also extended to the reviewers and the Editor for their valu-
able comments.
3353
References
Adamopoulos, G. I., & Pappis, G. P. (1996). A fuzzy linguistic approach
to a multicriteria-sequencing problem. European Journal of Operation
Research, 92, 628–636.
Beach, R., Muhlemann, A. P., Price, D. H. R., Paterson, A., & Sharp, J.
A. (2000). A review of manufacturing flexibility. European Journal of
Operational Research, 122, 41–57.
Bordogna, G., Fedrizzi, M., & Passi, G. (1997). A linguistic modeling of
consensus in group decision making based on OWA operators. IEEE
Transactions on Systems, Man and Cybernetics – Part A: Systems and
Humans, 27, 126–132.
Bordogna, G., & Passi, G. (1993). A fuzzy linguistic approach generalizing
boolean information retrieval: A model and its evaluation. Journal of
the American Society for Information Science, 4, 70–82.
Bresnen, M., Edelman, L., Newell, S., Scarbrough, H., & Swan, J. (2003).
Social practices and the management of knowledge in project environ-
ments. International Journal of Project Management, 21, 157–166.
Chuang, S. H. (2004). A resource-based perspective on knowledge
management capability and competitive advantage: An empirical
investigation. Expert Systems with Applications, 27(3), 459–465.
Collinson, S. (2001). Knowledge management capabilities in R&D: A
UK–Japan company comparison. R&D Management, 31(3), 335–347.
Desouza, K. (2003). Knowledge management don’t just manage crises.
New Zealand Management, 51.
Nonaka, I., & Takeuchi, H. (1995). The knowledge creating company.
Oxford: Oxford University Press.
Gerwin, D. (1993). Manufacturing flexibility: A strategic perspective.
Management Science, 39(4), 395–410.
Gold, A. H., Malhotra, A., & Segars, A. H. (2001). Knowledge
management: An organizational capabilities perspective. Journal of
Management Information Systems, 18(1), 185–214.
Grant, R. M. (1996). Prospering in dynamically competitive environ-
ments: Organizational capability as knowledge integration. Organiza-
tional Science, 7(4), 375–387.
Herrera, F., & Herrera-Viedma, E. (2000). Linguistic decision analysis:
Steps for solving decision problems under linguistic information. Fuzzy
Sets and Systems, 115, 67–82.
Herrera, F., Herrera-Viedma, E., & Verdegay, J. L. (1996). Direct
approach processes in group decision making using linguistic OWA
operators. Fuzzy Sets and Systems, 79, 175–190.
Herrera, F., Herrera-Viedma, E., & Verdegay, J. L. (1998). Choice
processes for non-homogeneous group decision making in linguistic
setting. Fuzzy Sets and Systems, 94, 287–308.
Herrera, F., & Martinez, L. (2000). A 2-tuple fuzzy linguistic represen-
tation model for computing with words. IEEE Transactions on Fuzzy
Systems, 8, 746–752.
Herrera, F., & Mart±́nez, L. (2001). A model based on linguistic 2-tuples
for dealing with multigranularity hierarchical linguistic contexts in
multiexpert decision-making. IEEE Transactions on Systems, Man and
Cybernetics – Part B: Cybernetics, 31(2), 227–234.
Johannessen, J., & Olsen, B. (2003). Knowledge management and sustain-
able competitive advantages: The impact of dynamic contextual
training. International Journal of Information Management, 23, 277–289.
Kacprzyk, J. (1986). Group decision making with a fuzzy linguistic
majority. Fuzzy Sets and Systems, 18, 105–118.
Kelly, D., & Amburgey, T. (1991). Organizational inertia and momentum:
A dynamic model of strategic change. Academy of Management
Journal, 34(3), 591–612.
Law, K. S., Wong, C., & Mobley, W. H. (1998). Toward a taxonomy of
multidimensional constructs. Academy of Management Journal, 41(6),
741–753.
Leonard, D. (1995). Wellsprings of knowledge: Building and sustaining the
source of innovation. Boston: Harvard Business School Press.
Liu, P. L., Wen, C. C., & Tsai, C. H. (2004). An empirical study on the
correlation between knowledge management capability and competi-
tiveness in Taiwan’s industries. Technovation, 24(12), 971–977.
3354 Z.-P. Fan et al. / Expert Systems with Applications 36 (2009) 3346–3354
Lubit, R. (2001). Tacit knowledge and knowledge management: The keys
to sustainable competitive advantage. Organizational Dynamics, 29(3),
164–178.
Torra, V. (1997). The weighted OWA operator. International Journal of
Intelligent Systems, 12, 153–166.
Vokurka, R. J., & O’Leary-Kelly, S. W. (2000). A review of empirical
research on manufacturing flexibility. Journal of Operations Manage-
ment, 18, 485–501.
Wang, R. C., & Chuu, S. J. (2004). Group decision-making using a fuzzy
linguistic approach for evaluating the flexibility in a manufacturing
system. European Journal of Operational Research, 154(3), 563–572.
Zadeh, L. A. (1965). A. fuzzy set. Information and Control, 8, 338–353.
Zadeh, L. A. (1975a). The concept of a linguistic variable and its
application to approximate reasoning (Part I). Information Science, 8,
199–249.
Zadeh, L. A. (1975b). The concept of a linguistic variable and its
application to approximate reasoning (Part II). Information Science, 8,
301–357.
Zadeh, L. A. (1976). The concept of a linguistic variable and its
application to approximate reasoning (Part III). Information Science,
9, 43–80.