Professional Documents
Culture Documents
Knowledge-Based Systems: Gin-Shuh Liang, Ji-Feng Ding, Chun-Kai Wang
Knowledge-Based Systems: Gin-Shuh Liang, Ji-Feng Ding, Chun-Kai Wang
Knowledge-Based Systems
journal homepage: www.elsevier.com/locate/knosys
a r t i c l e i n f o a b s t r a c t
Article history: The purpose of this paper is to apply a fuzzy quality function deployment (QFD) approach to prioritize
Received 14 August 2011 knowledge management (KM) solutions for an international port in Taiwan. First, the paper examines
Received in revised form 8 March 2012 house of quality (HOQ) matrices to facilitate handling of the ‘what’ (i.e., KM requirements) and ‘how’
Accepted 10 March 2012
(KM solutions) aspects of the QFD problem, and proposes procedures for the use of a fuzzy QFD method.
Available online 17 March 2012
A case study concerning port K in Taiwan is then used to demonstrate a systematic appraisal process for
prioritizing KM solutions, and twenty attributes with sixteen feasible KM implementation solutions are
Keywords:
measured employing an HOQ matrix. Finally, the top five feasible solutions for implementing KM at port
Knowledge management
Port
K are identified. The empirical results show that ‘establishment of a data storage and data mining system’
Fuzzy number in the technology dimension is the most urgent requirement for KM implementation at port K in Taiwan.
Quality function deployment Ó 2012 Elsevier B.V. All rights reserved.
Taiwan
0950-7051/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.knosys.2012.03.012
84 G.-S. Liang et al. / Knowledge-Based Systems 33 (2012) 83–91
improve efficiency, and lower operating costs, organizations numbers and the algebraic operations, linguistic values, and the
should treat their knowledge as an asset and manage it accord- ranking of fuzzy numbers, respectively.
ingly, share and transfer correct know-how, and transform
themselves into knowledge-oriented organizations. 2.1. Triangular fuzzy numbers and the algebraic operations
KM involves the use of a series of knowledge creation, acquisi-
tion, and application processes for the improvement of organiza- In a universe of discourse X, a fuzzy subset A of X is defined by a
tional performance. Hence, the success of an enterprise requires membership function fA(x), which maps each element x in X to a
both innovation and application of knowledge to gain competitive real number in the interval [0, 1]. The function value fA(x) repre-
advantages and promote KM [4]. A growing range of organizations sents the grade of membership of x in A.
are thus focusing attention on KM. In short, knowledge manage- A fuzzy number A [19] in real line is a triangular fuzzy number if
ment treats knowledge as an asset and manages it in a systematic its membership function fA : R ! ½0; 1 is
way to achieve the goal of enhancement of organizational perfor- 8
mance and competitiveness. < ðx cÞ=ða cÞ; c 6 x 6 a
>
In recent years, a growing body of literature [2,4,8,10– fA ðxÞ ¼ ðx bÞ=ða bÞ; a 6 x 6 b ð1Þ
>
:
12,20,21,26–28,31,40,41,45,46,48] has addressed the subject of 0; otherwise
KM, which underscores the widespread attention this topic has
received. However, a review of relevant literature suggests that with 1 < c 6 a 6 b < 1. The triangular fuzzy number can be de-
the application of KM is still limited to manufacturing, high-tech noted by (c, a, b). The parameter a gives the maximum grade of
industry, knowledge-intensive industry, service industries having fA(x), i.e., fA(a) = 1; it is the most probable value of the evaluation
frequent interactions with customers, and some government agen- data. In addition, ‘c’ and ‘b’ are the lower and upper bounds of the
cies. Despite the fact that there have been studies on the applica- available area for the evaluation data. They are used to reflect the
tion of KM to the maritime transport industry [47], discussion of fuzziness of the evaluation data. The narrower the interval [c, b],
its application to port management remains scanty. the lower is the degree of fuzziness of the evaluation data.
The integration of port service functions and their connection Let A1 = (c1, a1, b1) and A2 = (c2, a2, b2) be fuzzy numbers.
with the country’s industrial development not only result in bene- According to the extension principle [49], the algebraic operations
ficial economic clustering effects but also boost the market effec- of any two fuzzy numbers A1 and A2 can be expressed as
tiveness of the production sector. Moreover, the development of
global trade has further changed the management of traditional Fuzzy addition, :
maritime transport and led to critical changes in the international
maritime transport service market system. Haynes et al. [24] sug- A1 A2 ¼ ðc1 þ c2 ; a1 þ a2 ; b1 þ b2 Þ;
gested that factors connected with the success of a fourth-genera-
tion port operation [44] include technology, human resources, Fuzzy multiplication, :
resource integration strategies, and application of knowledge. With k A2 ¼ ðkc2 ; ka2 ; kb2 Þ; k 2 R; k P 0;
the development of KM and the knowledge-based economy, these
competitive port knowledge service capabilities have expanded to A1 A2 ffi ðc1 c2 ; a1 a2 ; b1 b2 Þ; c1 P 0; c2 P 0:
play a functional role, which further demonstrates the important
role played by KM in modern seaport operations.
2.2. Linguistic values
Ports are a type of service industry [3], but there are discrepan-
cies in perceived importance and satisfaction when KM is
In this paper, linguistic values [29,50] characterized by triangu-
implemented by port operators. In order to enhance KM imple-
lar fuzzy numbers are employed to express the fuzzy relationship
mentation performance, successful, concrete, feasibility solutions
degree between the KM requirements and solutions. According to
must be proposed in order to resolve the issue of these discrepan-
the practical needs and for matching the fuzzy QFD approach
cies. The quality function deployment (QFD) method offers just
developed in this paper, the fuzzy relationship degree set is defined
such a solution. There are situations in which information is
as S = {High, Medium, Low, Non}. The linguistic values can be
incomplete or imprecise, or when views subjective or endowed
defined as High = (0.5, 0.75, 1), Medium = (0.25, 0.5, 0.75),
with linguistic characteristics [50], creating a ‘fuzzy’ [49] deci-
Low = (0, 0.25, 0.5), and Non = (0, 0, 0), respectively.
sion-making environment. The fuzziness-based quality function
deployment (fuzzy QFD) approach can be used to evaluate the rela-
2.3. Ranking of fuzzy numbers
tionships between the requirements and solutions of implement-
ing KM for port operators. By using the fuzzy QFD approach, KM
In a fuzzy decision-making environment, ranking the alterna-
implementation requirements can be transformed into technical
requirements able to improve KM performance. tives under consideration is essential. To match the fuzzy QFD ap-
proach developed in this paper, and to solve the problem
In summary, the purpose of this paper is to apply the fuzzy QFD
method to prioritize technical KM solutions for an international powerfully, the graded mean integration representation (GMIR)
method, proposed by Chen and Hsieh [7] in 2000, is employed to
port in Taiwan. The first section provides background information.
The following section presents the research methodology, while determine the priority order of execution to all technical solutions.
Let Ai = (ci, ai, bi), i = 1, 2, . . . , n, be n triangular fuzzy numbers. By
the third section describes the fuzzy QFD procedures. The fourth
section consists of an empirical case study, and the final section the GMIR method, the GMIR value P(Ai) of Ai is
offers our conclusions. PðAi Þ ¼ ðci þ 4ai þ bi Þ=6 ð2Þ
Suppose P(Ai) and P(Aj) are the GMIR values of the triangular fuzzy
2. Research methodology numbers Ai and Aj, respectively. We define:
Ai > Aj $ PðAi Þ > PðAj Þ,
In this section, some of the concepts and methods used in this Ai < Aj () PðAi Þ < PðAj Þ,
paper are briefly introduced. These include the triangular fuzzy Ai ¼ Aj () PðAi Þ ¼ PðAj Þ.
G.-S. Liang et al. / Knowledge-Based Systems 33 (2012) 83–91 85
3. The fuzzy QFD approach 3.2. The systematic steps of fuzzy QFD approach
3.1. Basic concept of the QFD model Fuzzy QFD method implementation steps [15,29,30] are pro-
posed as follows.
The QFD model [13,14,23,25] can be used to translate customer
requirements into product specifications. It is a tool to employ the 3.2.1. Step 1. Identify the KM requirements
voice of the customer (VOC) in a search for the best product devel- Because KM requirements are not easy to obtain, we combined
opment solutions. Cohen [13] proposed a four-phase QFD model in determination of KM requirements via the academic literature
a discussion of product development; these phases respectively [2,5,6,9,15–17,32,34,42] with consultation with managers and ex-
consist of customer requirement planning (CRP), product charac- perts in Taiwan concerning the characteristics of KM implementa-
teristics deployment (PCD), process and quality control (PQC), tion at a port. Four dimensions with twenty KM requirement
and the operative instruction (OPI). In this paper, we focus on attributes for an international port are evaluated, and their codes
the CRP phase, which is used to develop procedures for identifying shown in parentheses.
KM requirements and solutions. The CRP phase of the QFD model
consists of use of a matrix, known as the ‘‘House of Quality’’ 1. Market competitiveness (R1). This dimension chiefly focuses on
(HOQ), which uses matrices to show multiple relationships be- attributes relating to the enhancement of competitiveness in
tween customer requirements (i.e., the ‘what’ requirements of the international port market, and includes the attributes of
KM) and technical specifications (i.e., ‘how’ the KM solutions have ‘improvement of high-level managers’ prediction ability (C1)’,
to be designed). In this paper, HOQ matrices are used to organize ‘solution of problems caused by business and change of envi-
KM requirements and establish the priority of solutions in order ronment (C2)’, ‘enhancement of knowledge sharing between
to meet KM requirements. In other words, the customer’s KM different departments (C3)’, ‘shortening the time needed for
requirements are transformed into technical specifications, which business procedures (C4)’, ‘improvement of employees’ work
are then evaluated to establish an impact ranking. Further discus- efficiency (C5)’, ‘strengthening of the port’s market penetration
sion and explanation of the QFD model may be found in Akao [1], ability (C6)’, and ‘reduction of port costs and expenses (C7)’.
Cohen [13], and Terninko [43]. 2. Human resources management (HRM) (R2). This dimension
A typical HOQ chart (American style) is shown in Fig. 1, and is concerns issues related to the improvement of HRM efficiency
created in six basic steps. The difference between the American- in port operation, and includes the attributes of ‘implementa-
style and Japanese-style HOQ charts is that the latter lacks Area E tion of a substitute system to enable any replacement employee
in Fig. 1. Since the Japanese style of HOQ is easy to use, however, to swiftly take over a new job in the event of a job transfer (C8)’,
it will be employed in this paper. ‘fast and effective solution of repeated, routine, and common
problems (C9)’, ‘standardization of working procedures to
Area A represents customer needs, as identified by the VOC. In improve employees’ business handling ability (C10)’, and
this paper, those needs are termed requirements, and influence ‘improvement of employees’ ability to apply information to
the KM solutions. The KM requirements are determined via a support decision-making (C11)’.
questionnaire. 3. Service innovation (R3). This dimension represents the ability to
Area B represents the relative importance of KM requirements, improve port service quality and organizational innovation
determined in this paper using the questionnaire. ability. Its attributes include ‘improvement of the innovation
Area C represents design requirements or technical specifica- abilities of the organization and its high-level managers
tions, which are ‘how’ the KM solutions to be made. In this through knowledge sharing (C12)’, enhancement of professional
paper, this ‘how’ question is the main issue, and involves iden- port managers’ success rate in their business negotiations with
tification of KM solutions. new customers (C13)’, ‘shortening of the time needed by the
Area D represents the relationship matrix, which is the core ele- port’s to respond to customer demands (C14)’, ‘improvement
ment of the HOQ. In this paper, the relationship strength is of the port’s online service quality and implementation of a
expressed by triangular fuzzy numbers. one-stop service system (C15)’, and ‘improvement of communi-
Area E represents the correlation matrix, which expresses how cation channels with the harbor bureau, shipping companies,
design requirements affect each other. Correlations are shown and shipping agents (C16)’.
with symbols, a rating scheme of 1–3–9, or linguistic 4. Organizational management (R4). This dimension focuses on
expressions. attributes associated with the improvement of a port’s organi-
Area F represents design requirement target values. In this zational management. It includes ‘improvement of organiza-
paper, the priorities of KM solutions are calculated. tional and individual knowledge learning ability (C17)’,
‘improvement of the return of unit labor cost (C18)’, ‘transform-
ing the port’s non-knowledge-oriented organization into a
knowledge-oriented organization and increasing knowledge
sharing within the organization (C19)’, and ‘application of KM
to the overall port operation management system to boost port
Correlation competitiveness (C20)’.
matrix (E)
3.2.2. Step 2. Comparing the importance and satisfaction levels of
Design requirements (C) attributes
The KM requirement attributes can be measured using a Likert-
Customer Relative importance of
type five-point scale [15,29] to evaluate the discrepancies between
requirements Relationship matrix (D) customer requirements
levels of perceived importance and satisfaction. If the latter are
(A) (B)
greater than the former, this implies that the port system KM
Target values (F) requirements are acceptable. Conversely, if the former exceed the
latter, this implies that some measures or solutions should be
Fig. 1. House of quality (HOQ). identified, and Step 3 executed. In this paper, the authors will eval-
86 G.-S. Liang et al. / Knowledge-Based Systems 33 (2012) 83–91
Table 1
Technical solutions of KM for an international port.
P
n P
n P
n
for each attribute (Ci) in Table 2. The rankings of the important
FSj ¼ W i cij ; W i aij ; W i bij ; j ¼ 1; 2; . . . ; m:
i¼1 i¼1 i¼1 KM requirements attributes for port K are then evaluated. The
top five KM requirement attributes were ‘reduction of port costs
and expenses (C7)’, ‘transforming the port’s non-knowledge-ori-
3.2.7. Step 7. Defuzzify the fuzzy relationship strength to rank the ented organization into a knowledge-oriented organization and
priority increasing knowledge sharing within the organization (C19)’,
The GMIR method, discussed in Section 2.3, is used to defuzzify ‘implementation of a substitute system to enable any replacement
the fuzzy weight FSj of technical solution Aj. The ranking value employee to swiftly take over a new job in the event of a job trans-
P(FSj) of FSj can be calculated by fer (C8)’, ‘enhancement of knowledge sharing between different
P
n P
n P
n departments (C3)’, and ‘improvement of the knowledge learning
PðFSj Þ ¼ W i cij þ 4 W i aij þ W i bij 6; j ¼ 1; 2; . . . ; m: abilities of individuals and the organization (C17)’, respectively.
i¼1 i¼1 i¼1
Second, the central part of Table 2 shows the fuzzy relationship
Based on these ranking values and the ranking rules in Sec- matrix. Four linguistic expressions were designed in the stage-two
tion 2.3, we can determine the priority of implementation of all questionnaire to measure the fuzzy relationship degree of each
technical KM solutions. attribute Ci corresponding to each solution Aj. Five valid question-
naires were collected to calculate the fuzzy relationship matrix
4. Case study [Rij]n m (see Step 5 in Section 3.2). After obtaining the fuzzy rela-
tionship matrix, the authors used the Steps 6 and 7 in Section 3.2
In this section, a case study of an international port (‘‘port K’’) in to calculate the fuzzy weights (FSj) and ranking values P(FSj) of
Taiwan is performed to demonstrate the calculation procedures all technical solutions. The technical solutions were ranked based
described above. on the ranking rules in Section 2.3, and these results are shown
at the bottom of Table 2.
4.1. Questionnaire and data collection The empirical results showed that the five best feasible solu-
tions for implementing KM at port K consist of ‘establishment of
In this paper, questionnaires for two stages were designed to a data storage and data mining system (A1)’, ‘establishment of a
evaluate the priority of KM requirements and the fuzzy relation- decision-making support system (A3)’, ‘construction of an informa-
ship degrees between KM requirements and technical solutions. tion and communication infrastructure (A2)’, ‘establishment of a
In the stage one questionnaire, we used twenty KM requirement database for document management (A5)’, and ‘use of groupware
implementation attributes to evaluate degrees of importance and and other software (A4)’, respectively.
satisfaction. One hundred personnel at port K, including senior
and middle managers and general unit personnel of port K, were 4.3. Discussions
selected to answer the survey questionnaires. A total of 74 valid
questionnaires (74%) were received from 100 respondents. The In summary, we have gained a rough idea concerning the tech-
survey was implemented through e-mails, phone calls, and in-per- nical requirements of KM implementation for port K, with atten-
son interviews by the authors. tion focused first on technology, followed by personnel, and then
The Cronbach’s a values of all twenty KM requirement imple- workflow. This implies that the development of various software
mentation attributes with respect to degrees of importance and and systems should first seek to structure data and information,
satisfaction were 0.9425 and 0.9375, respectively, showing that which should precede the application of KM awareness and abili-
the survey results had a satisfactory level of reliability [35]. The ties by port employees. Workflow-related tasks, such as system
content of the questionnaire was developed through an examina- and organizational reforms, should be performed during a later
tion of the academic literature, and investigated the characteristics stage.
of KM implementation at the port via the views of managers and Due to the fact that the environments faced by port operators
experts. A pretest was also conducted. As a result, the content can be very different, to ensure the correctness and timeliness of
validity of the questionnaire was considered very good. KM implementation strategies, the relationship between the
Regarding the stage-two questionnaire, the authors combined implementation of KM requirements and technical requirements
the twenty requirement attributes and sixteen technical KM should be constantly adjusted, so that they can serve as the basis
solutions to construct a matrix table evaluating fuzzy relationship for subsequent periodic evaluations of QFD methods. This paper
degree. Robbins [38] suggested five or seven decision-makers are has determined the top five feasible solutions for the implementa-
suitable for dealing with group decision-making problems. Due tion of quality technology deployment strategies, which are dis-
to the fact that the degree of fuzzy relationship is generated by a cussed as follows. These feasible solutions should be prioritized
professional committee, five experts consisting of senior managers, when port operators implement KM.
the majority of whom had been working in the field for over
twenty years, were selected to complete the stage two 1. Establishment of a data storage and data mining system. We
questionnaire. can see that information is obtained through the tabulation
and analysis of data, while knowledge is acquired through the
4.2. Empirical results classification and collation of information. After testing and
application, knowledge can be used to help enterprises make
The authors used the fuzzy QFD steps described in Section 3.2 to a profit, reduce costs, boost competitiveness and produce both
obtain the final results, shown in Table 2. The detailed construction tangible and intangible values. At the beginning of KM imple-
the HOQ in Table 2 is described below. mentation, port operators therefore need to first strengthen
First, the right side of Table 2 shows the xi ; yi ; wi , and Wi. The their ability to convert data into knowledge before their com-
Likert’s five-points scale [15,29] was designed in the stage-one petitiveness can be enhanced.
questionnaire to measure degree of importance and satisfaction, 2. Establishment of a decision-making support system. Research
and 74 valid questionnaires were collected to calculate the arith- on DSS has made great strides in recent years. The purpose of
metic means. The results showed that the arithmetic means of a DSS is to resolve uncertain decision-making problems faced
importance (xi ) are all larger than the degree of satisfaction (yi ) by high-level executives by integrating conventional data and
Table 2
88
The results of solutions of KM for international port K.
Solutions factors S1 S2
A1 A2 A3 A4 A5 A6 A7 A8 A9
R1
C1 (0.5, 0.75, 1) (0.5, 0.75, 1) (0.5, 0.75, 1) (0.375, 0.625, 0.875) (0.438, 0.688, 0.938) (0.5, 0.75, 1) (0.313, 0.563, 0.813) (0.313, 0.563, 0.813) (0.313, 0.563, 0.813)
C2 (0.375, 0.625, 0.875) (0.25, 0.5, 0.75) (0.438, 0.688, 0.938) (0.375, 0.625, 0.875) (0.125, 0.313, 0.5) (0.375, 0.625, 0.875) (0.313, 0.563, 0.813) (0.313, 0.563, 0.813) (0.375, 0.625, 0.875)
C3 (0.375, 0.625, 0.875) (0.375, 0.625, 0.875) (0.438, 0.688, 0.938) (0.438, 0.688, 0.938) (0.438, 0.688, 0.938) (0.375, 0.625, 0.875) (0.375, 0.625, 0.875) (0.438, 0.688, 0.938) (0.375, 0.625, 0.875)
C4 (0.438, 0.688, 0.938) (0.438, 0.688, 0.938) (0.438, 0.688, 0.938) (0.438, 0.688, 0.938) (0.5, 0.75, 1) (0.438, 0.688, 0.938) (0.375, 0.625, 0.875) (0.438, 0.688, 0.938) (0.375, 0.625, 0.875)
C5 (0.5, 0.75, 1) (0.5, 0.75, 1) (0.5, 0.75, 1) (0.438, 0.688, 0.938) (0.5, 0.75, 1) (0.438, 0.688, 0.938) (0.438, 0.688, 0.938) (0.5, 0.75, 1) (0.438, 0.688, 0.938)
C6 (0.25, 0.5, 0.75) (0.25, 0.5, 0.75) (0.125, 0.375, 0.625) (0.25, 0.5, 0.75) (0.313, 0.563, 0.813) (0.125, 0.375, 0.625) (0.188, 0.438, 0.688) (0.25, 0.5, 0.75) (0.188, 0.438, 0.688)
C7 (0.313, 0.5, 0.688) (0.313, 0.5, 0.688) (0.438, 0.688, 0.938) (0.313, 0.563, 0.813) (0.125, 0.313, 0.5) (0.375, 0.625, 0.875) (0.125, 0.313, 0.5) (0.125, 0.313, 0.5) (0, 0.188, 0.375)
R2
C8 (0.5, 0.75, 1) (0.375, 0.625, 0.875) (0.438, 0.688, 0.938) (0.5, 0.75, 1) (0.5, 0.75, 1) (0.375, 0.625, 0.875) (0.375, 0.625, 0.875) (0.188, 0.438, 0.688) (0.5, 0.75, 1)
C9 (0.5, 0.75, 1) (0.5, 0.75, 1) (0.5, 0.75, 1) (0.5, 0.75, 1) (0.5, 0.75, 1) (0.25, 0.438, 0.625) (0.438, 0.688, 0.938) (0.438, 0.688, 0.938) (0.5, 0.75, 1)
C10 (0.5, 0.75, 1) (0.5, 0.75, 1) (0.5, 0.75, 1) (0.375, 0.625, 0.875) (0.5, 0.75, 1) (0.438, 0.688, 0.938) (0.313, 0.563, 0.813) (0.25, 0.5, 0.75) (0.438, 0.688, 0.938)
C11 (0.25, 0.438, 0.625) (0.375, 0.563, 0.75) (0.125, 0.313, 0.5) (0.25, 0.438, 0.625) (0.438, 0.688, 0.938) (0.188, 0.438, 0.688) (0.25, 0.5, 0.75) (0.25, 0.5, 0.75) (0.438, 0.688, 0.938)
R3
(2)
(8)
0.0524
0.0498
0.0505
3.297
3.405
3.608
4.149
Note: The full names of all attributes of requirements and solutions of KM can be seen in the Section 3.2.
end, they also felt confidence they could apply this non-substitut-
GMIR values
managers can determine their KM capabilities to run their business methods to investigate patterns and advantages/disadvantages of
operations more efficiently and establish their core competence port KM implementation.
more effectively will be an important issue for future study.
References
[34] T.E. Notteboom, W. Winkelmans, Structural changes in logistics: how will port [43] J. Terninko, Step-by-Step QFD: Customer-Driven Product Design, St. Lucie
authorities face the change?, Maritime Policy and Management 28 (1) (2001) Press, Boca Raton, 1997.
71–89 [44] United Nations Conference on Trade and Development (UNCTAD), Technical
[35] J.C. Nunnally, I.H. Bersten, Psychometric Theory, McGraw-Hill, New York, Note: Fourth-Generation Port, UNCTAD Ports Newsletter 19 (1999) 9–12.
1994. [45] R. Vandaie, The role of organizational knowledge management in successful
[36] Organization for Economic Cooperation and Development (OECD), The ERP implementation projects, Knowledge-Based Systems 21 (8) (2008) 920–
Knowledge-Based Economy, Paris, 1996. <http://www.oecd.org/dataoecd/51/ 926.
8/1913021.pdf>. [46] W.P. Wagner, J. Otto, Q.B. Chung, Knowledge acquisition for expert systems in
[37] M. Polanyi, The Tacit Dimension, Anchor Books, New York, 1967. accounting and financial problem domains, Knowledge-Based Systems 15 (8)
[38] S.P. Robbins, Management, McGraw-Hill, New York, 1994. (2002) 439–447.
[39] P.M. Romer, Increasing returns and long-run growth, Journal of Political [47] C.C. Yang, P.B. Marlow, C.S. Lu, Knowledge management enablers in liner
Economy 94 (5) (1986) 1002–1037. shipping, Transportation Research Part E: Logistics and Transportation Review
[40] B. Rubenstein-Montano, J. Liebowitz, J. Buchwalter, D. McCaw, B. Newman, K. 45 (6) (2009) 893–903.
Rebeck, The knowledge management methodology team, A systems thinking [48] M.H. Zack, Developing a knowledge strategy, California Management Review
framework for knowledge management, Decision Support Systems 31 (1) 41 (3) (1999) 125–145.
(2001) 5–16. [49] L.A. Zadeh, Fuzzy sets, Information and Control 8 (3) (1965) 338–353.
[41] G. Shaw, A. Williams, Knowledge transfer and management in tourism [50] L.A. Zadeh, The concept of a linguistic variable and its application to
organisations: an emerging research agenda, Tourism Management 30 (3) approximate reasoning, Part 1, Information Sciences 8 (3) (1975) 199–249;
(2009) 325–335. L.A. Zadeh, The concept of a linguistic variable and its application to
[42] Y.L. Su, G.S. Liang, C.F. Liu, T.Y. Chou, A study on integrated port performance approximate reasoning, Part 2, Information Sciences 8 (4) (1975) 301–357;
comparison based on the concept of balanced scorecard, Journal of the Eastern L.A. Zadeh, The concept of a linguistic variable and its application to
Asia Society for Transportation Studies 5 (2003) 609–624. approximate reasoning, Part 3, Information Sciences 9 (1) (1976) 143–180.