Fuzzy Sets and Systems 63 0994) 131-139 North-Holland

131

A fuzzy multi-criteria decision making method for technology transfer strategy selection in biotechnology
Pao-Long Chang
h~stitnw of Management Science, National Chiao Tung University Taipei, Taiwan, R,O.C.

Yaw-Chu Chen
Instituw of Mamlgemetat Science, Natiomd Chiao Tung Universityand Department of International 7)'ade, Ming C/man College, Taipei, Taiwan, R,O,C
Received September 1993 Revised December 1993

Absmwc This paper discusses the potential application of fuzzy set theory to technology transfer strategy selection in the area of
biotcchnology management, An algorithm for technology transfer strategy selection is proposed. The algorithm is based on the concepts of fuzzy set theory and the hierarchical structure analysis. The linguistic variables and fuzzy numbers are used to aggregate the decision makers' subjective assessment about criteria weightings and appropriateness of alternative transfer strategies versus selection criteria to obtain the linal scores called fuzzy appropriateness indices. We also propose a revised method for ranking fuzzy numbers with index of optimism which was proposed by Kim and Park, Our revised method is based on the stage of data input for computing the total index of optimism in multi-person decision making prnblcm, instead of giving the index of optimism independently by a decision maker on the stage of information output, This new look at the index of optimism reflects the pooled risk-bearing attitude of multi-judge. The index of optimism is determined by the evaluation data conveyed by the decision makers at the beginning of the data input stage. Finally. we use the revised method to rank the fuzzy appropriateness indices for choosing the best technology transfer strategy.

Keywords: Multi-criteria decision making: technology transfer: index of optimism: ranking fuzzy nmnhe,'s.

1. Introduction Though application of fuzzy set theory in the fields of artificial intelligence, information processing, engineering, control, and human decision making has proliferated in recent years, effort in analyzing technology transfer strategy selection is lacking. Brown et al. [1] provided a technique that managers of government-sponsored R&D could use to evaluate innovations during their precommercialization stage in order to identify an appropriate technology transfer strategy. In their paper, the six types of successful commercialization strategies used by federal agencies were described: contracting R&D to industrial partners, working with industrial consortia, licensing to industry, influencing key decision makers, working with broker organizations, and generating end-user demand. Three sets of criteria for classifying innovations were proposed, including technological criteria, market criteria, and policy criteria. Technological criteria examine the innovations on the nature of science and technology. Market criteria evaluate innovations on the marketplace profiles, while policy criteria refer to the level of government support and the desired time
Correspondence to: Dr. Y.-C. Chen, Department of International Trade, Ming Chuan College, 250 Chuan-Shan N. Road, Section 5. Shihlin. Taipei, Taiwan,
0165-0114/94[$07.00 1994--Elsevier Science B.V. All rights reserved SSDI 0165-0114(94)00021-X

132

Pao-Ltmg Chmlg. Yaw.Ch, Chen I A .fuzzy multi-criteria decisiml makblg method

line. Then, they developed a guideline, which links the evaluation criteria and appropriate technology transfer strategy, for selecting a technology transfer strategy, based primarily on the five fully commercialized innovations. The Development Center for Biotechnology (DCB), partially supported by the government of Taiwan on a contract basis, is an autonomous nonprofit research and development organization established in March 1984 for the purpose of promoting and upgrading biotechnology industry in Taiwan. The functions of DCB are carried out along with two directions. Vertically, it will link up the academic sector and the industrial sector through developmental research. It has a pilot plant to develop technologies originated from academic researchers and then transfer them to appropriate manufactures. The Center also provides market research and assessment services for the academy, industry as well as the government. Horizontally, DCB will introduce, purchase and adopt suitable bioteehnologies from around the world with a view to transferring them to domestic industries. Biotechnology is a strategic industry in Taiwan. According to the current domestic technological capacity and industrial demand, the DCB has chosen pharmaceutical, agricultural, specialty chemicals and environmental protection as impending research task. However, DCB faces the problem of choosing the appropriate technology transfer strategy to insurc the innovation against unsuccessful commercialization. The transfer of technology from its source to commercial application is a very complex process. It is a multi-criteria decision making problcm in ill-structured situations. We must make a careful analysis among criteria, alternatives, weight, and decision makers before making a decision. Using the conventional crisp decision method, we always have to lind precise data. But under many conditions, we cannot get precise data because the data arc from the experience and the judgment of decision makers. For example, Brown et ai. define the evaluation criterion, 'nature of industry' of market criteria, with a crisp concept. They considered that an innovation under that criterion could be distinguished between 'competitive" and 'concentrated" in an exact term. And a ternary system (highly/sometimes/not) was used to represent the possible degrce of appropriateness of strategy versus criterion, Two observations arc in order at this point: (I) l~rs:ly, there is no sharp boundary between "competitive' and 'concentrated', The market data like econ~,nic data are imprecise and fuzzy by nature due to time, space, measurement error, or uncertain cnvironm~,,nt [7]. The use of conventional crisp approaches, both deterministic and random process, cut off information if the market concepts are intrinsically imprecise or fuzzy: (2) secondly, a technology transfer strategy versus the evaluation criteria, in some sense, should be taken into account with their admittedly imprecise degree of appropriateness, without too strict and unnatural approximating values. For example, a point system was given to the assessment of degree of appropriateness (highly appropriate = 2, sometimes appropriate = I, not--0) o[ a strategy versus a criterion [!]. The assessments of alternatives versus various criteria and the importance weight of criteria, however, often depend on judgment or approximation, Due to this type of existing fuzziness in decision making problem, we would expect a gradual transition from membership to nonmembership of an innovation's consonance with specified criteria as well as the relative importance weight of criteria. A revised method for ranking fuzzy numbers with index of optimism which was proposed by Kim and Park [ I I ] is also proposed in this paper. Our method is based on the stage of data input for computing the pooled risk-bearing attitude of multi-person decision making, unlike to the Kim and Park's method by giving the index of optimism unrelated to the evaluation data conveyed by the decision makers at the beginning of data input stage. We will use the revised method to rank the fuzzy numbers for technology transfer strategy selection. in this paper the triangular fuzzy numbers [9] and the linguistic values are utilized to assess the preference ratings of linguistic variables [14l. Section 2 discusses the definition of criteria and strategies. Section 3 deals with a fuzzy algorithm for technology transfer strategy selection. Section 4 gives a case study of hepatitis B vaccine technology transfer strategy selection, whereas Section 5 is discussion and conclusion.

Pap-Long Chang, Yaw-Cha Chen / A fuzzy multi-criteria decision making method

133

2. Definition of criteria and strategies

With the help of senior managers in DCB and business units, by means of personal interviews, the respondents were asked to list the criteria and commonly used strategies that were successful in transferring government-sponsored innovations. A committee of four decision makers, i.e., Dr, D2, D3, and /)4, has been formed to determine the most appropriate technology transfer strategy. After screening, four selection criteria are considered: Technological availability (Ct), Market potential (C2), Policy support (C3), and Management ability (C4). These criteria are described below. (1) Technological availability. This criterion describes the technology R&D by reflecting its likely rate of diffusion, including the process of product innovation, generic research or applied research, simplicity or complexity, proprietary or nonproprietary, low or high technological uncertainty, desired time of transfer, and transfer cost. (2) Market potential. This criterion includes the considerations of the breadth of possible applications, competitive or concentrated market, size of market, and product life cycle. (3) Policy support. This refers to the level of government support, including tax incentives or subsidies, infrastructure projects required for years to implement the R&D. (4) Management ability. This reflects whether the managerial and business functions of the recipient (firm) is effective or not, including manufacturing capability, financial and human resources, and marketing skills. The four commonly used strategies that were successful in transferring government-sponsored R&D in the case studies are described as follows: A~ = Purchasing (firms just buy R&D from research bodies). A2 = Working with industrial partner (like a joint venture strategy, but do not set up an individual firm, both share (50-50) the required resources). A3 = Licensh~g (the licensee has rights to produce the market the product). A4 = Cooperative R & D (the company's cost sharing are 100 percent; the company is committed to the commercialization process and as a way of enhancing the R&D effort).

3. A fuzzy algorithm for technology transfer strategy selection

The concepts of hierarchical structure analysis with two distinct levels are used in this paper. The first level is to evaluate fuzzy importance of the decision criteria (including technology innovations, market potential, policy support, and management ability) and the second level is to assign ratings to various strategies under each evaluation criterion. The attitudes of decision makers are reflected in the data input stage. The triangular fuzzy number and the linguistic variable are the two main concepts used in this paper to assess the preference ratings of linguistic variables, 'importance' and 'appropriateness'. The decision makers can employ an assumed weighting set W = {Very Low, Low, Medium, High, Very High} to assess the relative importance of various criteria. And use the linguistic rating set S = {Very Poor, Poor, Fair, Good, Very Good} to evaluate the appropriateness of the alternatives versus various criteria. The membership functions of the linguistic values in the weighting set W and linguistic rating set S can be represented by the approximate reasoning of triangular fuzzy numbers, as shown in Table 1. If the decision maker does not agree with the assumed preference rating system mentioned above, he can give his own rating by using the triangular fuzzy number to show the individual perception of the linguistic variables, 'importance' and 'appropriateness'. The reason of using triangular fuzzy numbers is that it is intuitively easy to be used by the decision makers. A more general representation of multi-criteria decision making problem is introduced. Suppose there is a committee of n decision makers (D~, D2. . . . . D,,) who are responsible for assessing the appropriateness of m alternatives (At, A2 . . . . . A,,,) under each of k criteria (C,, C2 . . . . . Ck) as well as the importance weight of the criteria. Let S,.0 (i = 1, 2 . . . . . m; t --- 1, 2 . . . . . k;j = 1, 2 . . . . . n) be the

134

Pao,Long Chang, Yaw.Chu Chen / A fuzzy muhi.criteria decision making method

Table 1. Membcrship functions for linguistic values Linguistic values Fuzzy numbers (0,0,0.25) (0, 0.25, 0.5)

Very Low (VL) Very Poor (VP)

Low (L) Poor (P)

Medium (M) Fair (F)

High (H) Good (G)

(0,25.0.5, 0,75) (0.5,0,75. I) (0.75, I. I)

Very High (VH) Very Good (V(;)

rating assigned to alternative A~ by decision maker Dj under criterion C,. Let W,j be the weight given to Ct by decision maker Dj. The committee has to aggregate the rating S#j of n decision makers for each alternative A~ versus each criterion C, to obtain the rating S#. Each pooled S, can further be weighted by weight W, according to the relative importance of the k criteria. Then, the final score F, fuzzy appropriateness index, of alternative A~ can be obtained by aggregating S# and W, Finally, rank the final scores ~. (i = 1.2 . . . . . m) to obtain the most appropriate alternative. in sum, an algorithm of the multi-person multi-criteria technology transfer strategy selection with fuzzy set approach is given in the following. Step !. Form a committee of decision makers, then identify the selection criteria and list to prospective technology transfer alternative strategies. Step 2. Choose the appropriate preference ratings for the importance weight of the selection criteria and the preference ratings for alternatives versus criteria. Step 3. Evaluate the importance weight of each criterion and the fuzzy ratings for the appropriateness of alternatives versus criteria. Step 4. Aggregate the wcight of criteria to get the aggregated weight W,; pool the decision makers' opinions to get the aggregated fuzzy rating S# of alternative A~ under criterion C,; then aggregate S# and Wr with respect to each criterion to obtain the fuzzy appropriateness indices F~ for all alternatives. Step S. Calculate the ranking value U.r(F~) associated with each alternative's fuzzy appropriateness index ~. Step 6. Choose the technology transfer strategy with the maximal ranking value.
4. An application - numerical treatment of the problem

The following example discusses the hepatitis B vaccine technology transfer strategy selection through the observation and survey of DCB and a company of Taiwan.

Step !. From Section 2, we know that the four decision makers (Dj, D,,, D,~, and/)4) are based on the four selected criteria (Ct, C,.. C~. and C~) in choosing the most appropriate strategy among the four alternative strategies (At, A~.A3. and A4). The definitions of evaluation criteria and alternative strategies are described in Section 2. Step 2, The decision makers can use either the importance weighting set W and appropriateness ratings set S described in Section 3, i.e., W = {Very Low, Low, Medium, High, Very High}, S = {Very Poor, Poor, Fair, Good, Very Good}, or his own preference rating by directly giving the triangular fuzzy numbers.

Pao-Long Chang, Yaw-Chu Chert / A fi~zzy multi-criteria decision making method Table 2. The importance of the criteria Criteria Decision makers
DI C~ C., (0.25, 0.5, 0.6) (0.55, 0.8, I).9) D_, VH H D3 H M D4 (0.3, 0.4, 0.6) (0.5, 0.6, 0.8)

135

C,
C4

(0.75, 0.9. 1)
(0.55, 0.8.0.9)

H
M

VH
H

(0.5, 0.6, 0.85)

(0.4, 0.5, 0.7)

Table 3. The decision makers' evaluation of the four alternatives under criterion Ct Alternatives Decision makers

Di
At Aa A~ A4 ((),55, 1|,8, 0.9) ([). I, 0.3, 0,4) (0. I. [).3, 0.4) (0,25, [I.5.0.6)

D.,
VG F P F

D~
G P P P

D4
(0.5, 0.6, [).8) (0, 0.2, 0.5) ([).3, 0.4, 0.6) (0. I, 0.2, 0.35)

Step 3. The weight assigned to the four criteria given by the decision makers is presented in Table 2. The appropriateness of alternatives versus criteria given by decision makers are presented in Tables 3-6. Step 4. Many methods have been proposed to pool the decision makers' opinions, for example, mean, median, max, min, and mixed operators [3]. Each of operators has its own limitations. Criteria for selecting appropriate aggregation operator can be found in Zimmermann [15]. Since the average
Table 4. The decision makers' evaluation of the four alternatives under criterion C, Alternatives Decision makers
Di D: D~ D4

A~ A., A.~ A4

(0.25, 0.5, O.fi) ((I.25, 0.5, 0,6) ((I.55, 0.8, 0.9) (0.55.0.8, 0.9)

VP P C; VG

P P G G

((). I, 0.2, 0.35) (0.5, 0.6, 0.8) (0.5, 0.8, 0.85) (0.7, 0.8, 0.9)

Table 5. The decision makers" evaluation of the four alternatives under criterion C~ Alternatives Decision makers Di AI A-, A~ A4 (0.2.0.3.0.5) (0.0.2.0.3) (0.25, 0.5.0.6) (0.1,0.3.0.4) D, VP P F F D3 P VP G G D,~ (0. I. 0.2, 0.3) (0.3,0.4,0.6) (0.5, 0.6, 0.85) (0.2,0.3,0.5)

136

Pao,Long Chang. Yaw.Chu Chen/ A fi~zzy multi-criteriadecision making method

"[able 6. The decision makers" evaluation of the four alternatives under criterion C,~ Alternatives Decision makers

Di
A~ A_, A~
,4.4

D..
VG F F
G

D3
G G F
P

D+
(0.5, 0.6, 0.85) (0.5, 0.7, 0.9) (0.3, 0.4, 0.6)
(0.3, 0,4, 0.55)

(0.55.0,8, 0,9) (0,25.0.5, 0.6) (0.25.0.5, 0.6)

(0.25.0.5, 0.6)

operation is the most commonly used aggregation method, we will use the mean operator to aggregate the decision makers assessments. Let (D and be fuzzy addition and multiplication operators. Define

S,, = ( ~ ) ( S,, , (D S,, (D . . . (D S,,,, )

and

(i)

w, = (~) (w,, e w,:e. . q~ w,,,)

(2)

where S,, is the average fuzzy appropriater~ess rating of alternative A~ under criterion C, and Wt is the average importance weight of criterion C,. Thus, the fuzzy appropriateness index F~ of the ith alternative can be obtained by aggregating S# and W,. denoted as

g = (~) l(S,, w,) ~(s,: w.4 ~ . . . (~ (s,~ w~)].

(3)

Let S+,~= (q#i. o , , p,~) and W,j = (c,n ++,nb,i) be triangular fuzzy numbers. Then F~ can be approximately obtained by

F)m(Y,,Q,.Zo)
where
/,

(4)

y = ~ q,,c, tlk-.
I tl |

O, = ~ ,,,,,,,/k,
t I #l

z, = ~, p,,b,/k,
t, | ~

q,, = y_, ,~,,,/,,,

i:l

&)

o.

=~
i'1
#0

o,,In.

P., = ~, p,,,I,,
for i = 1.2 . . . . .

c, =

c,i/n,
..... k:j=l.2

a, = ~, a,Jn.
..... n.

b, = ~.b,,I..
l-!

m: t = l . 2

Aggregating the S+, and 14',by using (4). we obtain the fuzzy appropriateness indices (F~) as shown in Table 7.
Table 7. The fuzzy approprialcness four alternatives
Allernalives A) A,

indices of Ihe

Fuzzy appropriatuness indices F~ = (I).1473, 0.3334, 0,5662) /-~ = (0.0830, 0.2588, 0.5063) ~ = (0,1554, 0.373 I, 0.6360) F~ = (0.1561, 0,3684, 0,6130)

A,
'!4

Pao-Long Chang, Yaw-Chu Chen / A fiezzy multi-criteria decision making method

137

S t e p 5. There are many metkods for ranking fuzzy numbers [2, 4-6, 8, 10, 11]. In this paper, we will modify the Kim and Park [1i] method to compute the ranking values of fuzzy appropriateness indices under a group of decision makers. Let F~ (i = 1, 2 . . . . . m) be the fuzzy appropriateness indices of m alternatives. The corresponding membership functions can be obtained by (4). Define the maximizing set M = {(x, At(x)) [ x ~ R} with

xL

otherwise, and minimizing set G = { ( x , f ~ ( x ) ) [ x

x,

~ R } with

f~;(x) = 0,

otherwise,

where xt = infS, x., = sup S, S = U['~l F,, E = {x Ifi~(x) >0}, for i = 1, 2 . . . . . m. Define the optimistic utility UM(E) and pessimistic utility U,(F,) of each fuzzy appropriateness index E as
UM(5) = sup (5)

and U~(Fj) = 1 - sup (fFa(X) ^ f c ; ( X ) )

x

(6)

for i = 1, 2 . . . . . m , w h e r e ^ means min. Define ranking value U.r(F~) of fuzzy appropriateness indices as
U , . ( F J = aUM(F ) + (l - c )U.(FJ, o, <<-1.

(7)

The value a is an index of rating attitude. It reflects the decision maker's risk-bearing attilud. However, it is difficult to apply the Kim and Park's method directly in multi-person decision making problem. According to their method, we first need to determine an index of optimism before ranking the fuzzy appropriateness indices of alternatives. It is reasonable to evaluate the index of optimism through the evaluation data conveyed by the decision makers. It is unlikely in the way of Kim and Park that the index of optimism is given by a decision maker at the data output stage. A method to determine the index of optimism in multi-person decision making is proposed as follows: Let B = (c, a, b) be a normal triangular fuzzy number. It may be the evaluation of the importance of various criteria or the appropriateness of alternatives versus various criteria. Define Y = (a - c ) / ( b - c ) as the index of rating attitude of an individual decision maker. It reflects the decision maker's risk-bearing attitude. If 1' > 0.5, it implies that the decision maker is a risk lover. If y < 0.5, the decision maker is a risk avertor. If Y = 0.5, the attitude of decision maker is neutral to the risk. Thus, the total index of rating attitude,/3, with the evaluation data of individuals can be obtained by
[3 =
j=l

(a,j - c q ) / ( b q - co) +

~
i=l t=l j=l

(o,j - q , j ) / ( P i ~ - q , j ) / ( k x n + m x k x n )

(8)

where Sitj = (qio, oitj, P~o) and W,j = (c,j, a o, btj) described as mentioned above, for i = 1, 2 . . . . . m: t = 1, 2 . . . . . k: j = 1, 2 . . . . . n. By using (4)-(6) and (8), the ranking values Ur(F,) can be approximately obtained by
U-r(F~) _m /3[(Zi - x , ) [ ( x 2 - x , - Qi + Z i ) + (1 -/3)[1 - (x2 - Y~)/(x2 - xt + Q, - Y,)]

(9)

138

Pao-Long Chang, Yaw.Chu Chen / A filzzy multi-criteria decision makb~g method

Table 8. The ranking values of the fuzzy appropriateness indices for alternatives Alternatives Ai A2 A3 A4 Ranking values 0.4852 0.3937 0.5362 0.5275

for i = 1, 2. . . . . m: where x, = min{t, ~ . . . . . E,,} and x2 = max{Zt, Z2 . . . . . Z,,,} and/3 is expressed in (8). We know that 3' and/3 vary in [0, 1]. By using (9) we obtain the ranking values of E as shown in Table 8. Step 6. The ranking order of fuzzy appropriateness indices for four alternatives is Ur(F.~), Ur(F4), Ur(F~), and Ur(~). Therefore, it is obvious that the best selection of technology transfer strategy is A~, i.e. the licensing strategy. Now, the committee can recommend that the alternative A3 is the most appropriate technology transfer strategy.

$. Discu~ion and conclusion

In this paper, a decision algorithm based on the fuzzy set theory is proposed to solve the technology transfer strategy selection problem. This paper also suggests a new look on the concept of the index of optimism that is mentioned by Kim and Park [1 I]. in Section 4, we give the definitions of 3, and fl for computing the individual decision maker's and group decision makers' risk-bearing attitude respectively. From Tables 2-6, we can also obtain ~,~ =0.68, -/:=0.55, 3'3=0.5, 3'4 = 0.3907, and /3 = 0.5555302. The result implies that D, and D: are risk lovers. D.~ is neutral to the risk. While/)4 is a risk avertor. It also reflects that on the whole the committee of four decision makers is a risk lover, because/3 is greater than 0.5. Therefore, it is reasonable to evaluate the risk-bearing attitude based on the evaluation data conveyed by the decision makers at the initial stage of data input, but not on the data output stage. The concepts of linguistic variable and fuzzy number are also used in this paper, because they can easily bc used to describe the subjective assessments of the appropriateness of alternatives versus criteria and the importance weightings of criteria. There are many research issues remaining in the development of this approach. Among these are the following: (!) The definition of the appropriate fuzzy linguistic variables, their numbers, their values, and their universe of discourse for a general use in the algorithm. (2) A procedure for collecting raw data and estimating the proper membership functions for the base linguistic variable values. (3) The criteria tree of technological availability, market potential, policy support, and management ability can be decomposed into more levels of hierarchical structure, i.e., down to the subcriteria. Then the evaluation task is performed on the level of subcriteria. The decision algorithm proposed in this paper can be also used to find the best technology transfer strategy. (4) The relative importance of every decision maker is not considered in this paper. In many multi-judge decision problems, the weights Wtj are not only determined by decision maker j and criterion t, but also can be varied by the decision environment and structure [12]. Although the decision algorithm in this paper is primarily designed for the technology transfer strategy selection, it can be applied to problems such as project management, foreign market entry mode selection, and many other areas of management decision problems.

Pao-Long Chang. Yaw-Chu Chen / A .hlzzy multi-criteria decision making method

139

References
[1] M.A. Brown, L.G. Berry and R.K. Goel, Guidelines for successfully transferring government-sponsored innovations, Research Policy 20 (1991) 121-143, [2] G. Bortolan and R. Degani, A review of some methods for ranking fuzzy subsets, Fuzzy Sets and SystenTs 15 (1985) 1-19, [3] J.J. Buckley, The multiple judge, multiple criteria ranking problem: A fuzzy set approach, Fuzzy Sets and S.vstems 13 (1984) 25-37. [41 J.J. Buckley, Ranking alternatives using fuzzy numbers, Fuzzy Sets and Systems 15 (1985) 21-31. [5] J.J. Buckley and S. Chanas, A fast method of ranking alternative using fuzzy numbers, Fuzzy Sets and Systems 30 (1989) 337-338. [6] L.M. Campos Ib~ifiezand A. Gonzalez Mufioz, A subjective approach for ranking fuzzy numbers, Fuzzy Sets and Systems 29 (1989) 145-153. [7] P.L. Chang and Y.C. Chen, Fuzzy number in business conditions monitoring indicators: fuzzy set methodologies in economic condition, in: Fuzzy Engineering reward Human Friendly Systems (Proceedings of the International Fuzzy Engineering Symposium '91, Yokohama, Japan) 2 (1991) 11191-1100. 181 S,H. Chert, Ranking fuzzy numbers with maximizing set and minimizing set, Fuzzy Sets and Systems r/(1985) 113-1311. [91 D. Duhois and H. Prade, Operations on fuzzy numbers, htt,,rnathmal ,hmrnal of Systems Science 9 (1978) 613-629. 110] A. Gonaz~ilez Muit~, A study of the ranking function approach through mean wdues, l:,zzy Sets" and Systenls 35 (19911) 29-41. I111 K. Kim and K.S. Park, Ranking fuzzy numbers with index of optimism, Fuzzy Sets and Sysu,ms 35 (199{)) 143-1511. 112] P.Z. Wang, D.Z. Zhung, H.M. Zhang ,'rod K.C. Yau, Degree analysis and its application in decision making, Proceeding of the Second btternational Workshop on ArtiJicial bttelligence in Economics arid Management. Singapore (1989). [13] L.A. Zadch, Fuzzy sets, Information and Control 8 (1965) 338-353. 114] L.A. Zadeh, The concept of a linguistic variable and its application to approximate reasoning, hl.hmnation Science 8 (1975) 199-249 (1), 301-357 (!!); 9 (1976) 43-80 (III). [15] H.-J. Zimmermann, Fnzzy Sets, Decision Makb~g, and Expert Systems (Kluwer Academic Publishers, Boston, 1987).