The Evaluation of rock bands using an Integrated MCDM Model - An Empirical Study based on Finland (2000s)

Transcription

1 The Evaluation of rock bands using an Integrated MCDM Model - An Epirical Study based on Finland (2000s) a b Sarfaraz Hashekhani Zolfani a Meysa Donyavi Rad b Sa.hashekhani@gail.co eysa.donyavirad@yahoo.co Departent of Industrial Engineering, Shoal University, Aol, Iran Departent of Industrial Manageent, Kar Institute of higher education, Qazvin, Iran Abstract Selecting and ranking usicians and rock bands always is so hard and doesn t have a specific ethod for evaluation. The ai of this research is perforance evaluating of the rock bands. One of the liitations of this study is priary selection of ulti-role artists and this proble was solved by using the Delphi ethod. Criteria of this study identify fro group of experts that consist of: Technical Ability, Coposing, Moral, Behavior and lyrics, Ideology and Innovation, Vocalist. In this study finish rock bands selected for evaluating in 2000s. Copetitive world causes bands cant copose a high level albu ust in a one year and usually take tie to copose a high level albu. It s so iportant to consider that bands also participate is tours because of this it sees good to evaluating bands in a decade and another advantage is results can be consider in recent years. Final selected rock bands obtain fro experts also: The Rasus, H.I.M, Nightwish. Two MCDM ethods are used in the evaluating process. Fuzzy AHP is applied to deterine the relative weights of evaluation criteria and VIKOR is applied for evaluating and ranking final selected rock bands.this study can be used as a pattern for evaluating other groups of usic artists and countries and also can be used for awards. Keywords: Rock Bands, Perforance Evaluating, Fuzzy AHP, VIKOR 1. Introduction Music is not erely a for of entertainent but also the easiest way of counication aong people, a ediu to share eotions and a place to keep eotions and eories. Music can express eotion in succinctly but in an effective way. Peoples select different usic at different tie concordance with listening tie s ood and obectives (E. E. PeMyint and M. Pwint, 2010). Musical genres, the highlevel descriptor of usic which are created and used by huans for categorizing and describing the vast universe of usic are extensively used in usic stores, radio stations, and the Internet (D. Jang et all, 2008). Rock usic is a genre of popular usic that developed during and after the 1960s. It has its roots in 1940s and 1950s rock and roll, rhyth and blues, country usic and also drew on folk usic, azz and classical usic. The sound of rock often revolves around the electric guitar, bass guitar, drus, and keyboard instruents such as Haond organ, piano, or, since the late 60s, synthesizers. Rock usic typically uses siple rhyths in a 4/4 eter, with a repetitive snare dru back beat on beats two and four (Hashekhani Zolfani et all, 2011). In the late 1960s and early 1970s, rock usic developed different subgenres. When it was blended with folk usic it created folk rock, with blues to create blues-rock and with azz, to create azz-rock fusion. In the 1970s, rock incorporated influences fro soul, funk, and Latin usic. Also in the 1970s, rock developed a large nuber of subgenres, such

2 as soft rock, gla rock, heavy etal, hard rock, progressive rock, and punk rock. Rock subgenres that eerged in the 1980s included new wave, hardcore punk and alternative rock. In the 1990s, rock subgenres included grunge, Britpop, indie rock, and nu etal. A group of usicians specializing in rock usic is called a rock band or rock group (Hashekhani Zolfani et all, 2011). Many rock groups consist of an electric guitarist, lead singer, bass guitarist, and a druer, foring a quartet. Soe groups oit one or ore of these roles or utilize a lead singer who plays an instruent while singing, soeties foring a trio or duo; others include additional usicians such as one or two rhyth guitarists or a keyboardist. Rock bands fro soe genres, particularly those related to rock's foundations in rock and roll, include a saxophone. More rarely, groups also utilize bowed stringed instruents such as violins or cellos, and brass instruents such as trupets or trobones. In the new illenniu the ter rock has been used as a blanket ter including fors such as pop usic, reggae usic, soul usic, and soeties even hip hop, with which it has often been contrasted through uch of its history (R. Beebe et all, 2002). Rock usic is the ost popular cultural phenoenon of the second half of the twentieth century and the single greatest propagator of the oral, social, and religious values of our society. Social analysts concur that rock usic has becoe a priary force in shaping the thinking and life-style of this generation (Bacchiocchi, 1999). The worldwide popularity of rock usic eant that it becae a aor influence on culture, fashion and social attitudes. Different sub-genres of rock were adopted by, and becae central to, the identity of a large nuber of subcultures. In the 1950s and 1960s, respectively, British youths adopted the Teddy Boy and Rockers subcultures, which revolved around US rock and roll. The counter-culture of the 1960s was closely associated with psychedelic rock (M. Brake, 1990). In the past decade rock genre had a professional change and experience and any young boys and girls have liked to be a rock star and a professional usician in rock genre. We can see people all around world who like to listen to rock bands and rock superstars (Hashekhani Zolfani et all, 2011). This paper will use new ethods for selecting the best rock band and create a fraework for future research and awards like Gray awards. There isn t any award for bands and artists that evaluated in a decade. Finland has good potential in usic and especially in rock genre that has any fans fro all over the world. The purpose of this paper is to select the best rock band in 2000s in Finland as a case study. 2. Research Method 2.1. Delphi Method The Delphi Method was first developed by Dalkey and Heler (1963) in corporation and has been widely applied in any anageent areas, e.g. forecasting, public policy analysis and proect planning. However, the traditional Delphi Method also has soe disadvantages, such as low convergence expert opinions, high execution cost, the possibility of filtering out particular expert opinions. A typical Delphi survey approach consists of 3 rounds of questionnaires that respondents consecutively answer in a tiely fashion (1). The Delphi ethod accuulates and analyzes the results of anonyous experts that counicate in written, discussion and feedback forats on a particular topic. Anonyous experts share knowledge skills, expertise and opinions until a utual consensus is achieved (Murry& Haons, 1995). The Delphi ethod consists of five procedures: (a) select the anonyous experts; (b) conduct the first round of a survey; (c) conduct the second round of a questionnaire survey; (d) conduct the third round of a questionnaire survey; and (e) integrate expert opinions and to reach a consensus. Steps (c) and (d) are norally repeated until a consensus is reached on a particular topic (Murry& Haons, 1995). The Delphi ethod has proven a popular tool in inforation systes (IS) research [7,8,9]. Citing a lack of a definitive ethod for conducting the research and a lack of statistical support for the

3 conclusions drawn, Schidt [10] presented a step-wise ethodology for conducting such studies. The Delphi ethod originated in a series of studies that the RAND Corporation conducted in the 1950s. The obective was to develop a technique to obtain the ost reliable consensus of a group of experts. While researchers have developed variations of the ethod since its introduction, Linstone and Turoff [11] captured coon characteristics in this description: Delphi ay be characterized as a ethod for structuring a group counication process so that the process is effective in allowing a group of individuals, as a whole, to deal with a coplex proble. To accoplish this structured counication there is provided: soe feedback of individual contributions of inforation and knowledge; soe assessent of the group udgent or view; soe opportunity for individuals to revise views; and soe degree of anonyity for the individual responses. Delphi researchers eploy this ethod priarily in cases where udgental inforation is indispensable, and typically use a series of questionnaires interspersed with controlled opinion feedback [12]. Researchers have applied the Delphi ethod to a wide variety of situations as a tool for expert proble solving. One variant that has received widespread use is the ranking-type Delphi, used to develop group consensus about the relative iportance of issues. Schidt provides a detailed description of how to conduct this type of Delphi survey, including guidelines for data collection, data analysis (based on non-paraetric statistical techniques), and reporting of results. Forecasting and issue identification/prioritization represent one type of application of the ethod. Concept/fraework developent represents a second type of application of the Delphi ethod [13]. Identifying experts: In alignent with the guidelines of Delbecqet all [14], the study will use a ultiple step iterative approach to identify the experts: Step 1: Prepare KRNW Identify relevant disciplines or skills : acadeics, practitioner,governent officials, officials of NGOs Identify relevant organizations Identify relevant acadeic and practitioner literature Step 2: Populate KRNW with naes Write in naes of individuals in relevant disciplines or skills Write in naes of individuals in relevant organizations Write in naes of individuals fro acadeics and practitioner literature Step 3: Noinate additional experts Contact experts listed in KRNW Ask contacts to noinate other experts Step 4: Rank experts Create four sub-lists, one for each discipline Categorize experts according to appropriate list Rank experts within each list based on their qualification Step 5: Invite experts Invite experts for each panel, with the panels corresponding to each discipline Invite experts in the order of their ranking within their discipline sublist Target size is Stop soliciting experts when each panel size is reached Fig.1. Procedure for selecting experts in the exaple study. *Prepare a knowledge resource noination worksheet (KRNW)

4 Table 1: saple knowledge resource noination worksheet Skills Acadeic Music Instructor Professional Musicians: Pianist Guitarist (Lead, Rhyth, Bass) Vocalist Coposer Literature: Acadeic Literature Instructor Song Writer Experts of usic (conteporary) Organization and Counity University Musicians & Rock Bands University & Song Writers 2.2. Selecting Experts For this research by considering the liitations 11acadeic usic instructor were selected fro University consist of: 6 PhD and 5 MSc and 35 persons fro professional usicians including : 5 pianist, 14 guitarist (7 lead guitars, 4 rhyth guitar, 3 bass guitar), 8 lead vocals and 8 coposer and 15 persons for evaluating lyrics including : 8 acadeic literature and 7 song writers and at the end 20 persons were selected as experts of conteporary usic Delphi Study Process: Phase 1: Brainstoring * For this phase only, treat experts as individuals, not panels * Questionnaire 1: Ask experts to list 6 best rock bands Phase 2: Process of selecting criteria * Questionnaire 3: for introduce and selection of criteria * Questionnaire 4: give feedback to experts to final selection of criteria Phase 3: Ranking * Give points to each selected bands in criteria that were selected in * Using VIKOR for ranking bands Figure 2: Delphi study adinistration process 2.4. Fuzzy AHP A good decision-aking odel needs to tolerate vagueness or abiguity because fuzziness and vagueness are coon characteristics in any decision-aking probles (Yu, 2002). Since decisionakers often provide uncertain answers rather than precise values, the transforation of qualitative preferences to point estiates ay not be sensible. Conventional AHP that requires the selection of arbitrary values in pairwise coparison ay not be sufficient and uncertainty should be considered in soe or all pairwise coparison values (Yu, 2002). Since the fuzzy linguistic approach can take the optiis/pessiis rating attitude of decision-akers into account, linguistic values, whose ebership functions are usually characterized by triangular fuzzy nubers, are recoended to assess preference ratings instead of conventional nuerical equivalence ethod (Liang & Wang, 1994). As a result, the fuzzy-ahp should be ore appropriate

5 and effective than conventional AHP in real practice where an uncertain pairwise coparison environent exists (Lee, Chen, & Chang, 2008).There are any fuzzy-ahp ethods proposed by various authors (Buckley, 1985; Chang, 1996; Cheng, 1997; Deng, 1999; Leung & Cao, 2000; Mikhailov, 2004; Van Laarhoven& Pedrycz,1983). These ethods are systeatic approaches to the alternative selection and ustification proble by using the concepts of fuzzy set theory and hierarchical structure analysis. Decision-akers usually find that it is ore confident to give interval udgents than fixed value udgents. This is because usually he/she is unable to explicit about his/her preferences due to the fuzzy nature of the coparison process. In this study, we prefer Chang (1996) extent analysis ethod because the steps of this approach are easier than the other fuzzy-ahp approaches (Fuzzy sets and AHP are not detailed here because of being well-known applications). The steps of Chang (1996) extent analysis approach are as follows: Let X = {x1,x2,...,xn} be an obect set, and U = {u1,u2,...,u} be a goal set. According to the ethod of Chang (1996) extent analysis, each obect is taken and extent analysis for each goal, gi, is perfored, respectively. Therefore, extent analysis values for each obect can be obtained, with the following signs (Dag deviren, Yuksel, & Kurt, in press): M 1 1 g1, M g1,, M 1 g1, i=1, 2,..., n, (1) Where all the M gi ( = 1,2,...,) are triangular fuzzy nubers.the steps of Chang s extent analysis can be given as in the following: Step 1: The value of fuzzy synthetic extent with respect to the ith obect is defined as n S i = =1 M gi [ i=1 =1 M gi ] 1 (2) To obtain =1 M gi, perfor the fuzzy addition operation of extent analysis values for a particular atrix such that M gi =1 =( =1 l, =1, =1 u i ). (3) n And to obtain [ i=1 =1 M gi ] 1, perfor the fuzzy edition operation of extent analysis values for a particular atrix such that n i=1 =1 M gi =( =1 l, =1, =1 u i ). (4) and then copute the inverse of the vector in Eq. (4) such that n [ i=1 =1 M gi ] 1 = ( 1 =1 l, 1 =1, 1 =1 u i ) (5) Step 2: The degree of possibility of M 2 = (l 2, 2, u 2 ) M 1 = (l 1, 1, u 1 ) is defined as V=(M 2 M 1 )=sup[in(μ M1 (x), μ M2 (y))] (6) and can be equivalently expressed as follows: V=(M 2 M 1 )=hgt(m 1 M 2 ) = μ M2 (d)

6 = { 1 if 2 2, 0 if l 2 u 2, l 1 u 2 ( 2 u 2 ) ( 1 l 1 ) otherwise, (7) Where d is the ordinate of the highest intersection point D between μ M1 and μ M1 (see Fig. 2). To coparem 1 and M 2, both the values of V=(M 2 M 1 ) and V=(M 2 M 1 ) are needed. Step 3: The degree possibility for a convex fuzzy nuber to be greater thank convex fuzzy nubers M i (i = 1,2,...,k) can be defined by V(M M 1,, M 2,, M k ) = V[(M M 1 ) and (M M 2 ) and... and (M M k )], = in V(M M i ), i = 1, 2,, k. (8) Assue that d / = (A i ) in V(S i S K ) (9) For k = 1,2,...,n ; k i. Then the weight factor is given by Fig 3: The intersection between M 1 and M 2. W / = (d / (A 1 ), d / (A 2 ),..., d / (A n )) T, where A i (i = 1,2,...,n) are n eleents. (10) Step 4: Via noralization, the noralized weight vectors are W / = (d(a 1 ), d(a 2 ),..., d(a n )) T, where W is a nonfuzzy nuber. 2.5.VIKOR METHOD Introduction to VIKOR The VIKOR ethod is a coproise MADM ethod, developed by Opricovic.S and Tzeng (Opricovic,1998;Opricovic, S. and Tzeng, G. H., 2002) started fro the for of Lp-etric :

7 L pi ={ n [w (f f i ) (f f )] p 1 p =1 } 1 p + ; i=1,2,,i. The VIKOR ethod can provide a axiu group utility for the aority and a iniu of an individual regret for the opponent (Opricovic,1998; Opricovic, S. and Tzeng, G. H., 2002 ; SerafiOpricovic and Gwo-Hshiung Tzeng,2004) VIKOR Steps 1) Calculate the noralized value f i = x i n 2 x i =1 i=1,2,.., ; =1,2,,n 2) Deterine the best and worst values For all the attribute functions the best value was f and theworst value was f that is, for attribute J =1- n, we get forulas (2) and (3) f =axf i, i=1,2,, (2) f =inf i, i=1,2,, (3) Where f is the positive ideal solution for the thcriteria,f is the negative ideal solution for the th criteria. If one associates all f one will have the optial cobination, which gets the highest scores, the sae as f. 3) Deterine the weights of attributes The weights of attribute should be calculated to express their relative iportance. 4) Copute the distance of alternatives to ideal solution This step is to calculate the distance fro each alternative to the positive ideal solution and then get the su to obtain the final value according to forula (4) and (5). n S i = =1 w (f f i )/(f f ) (4) R i = Max [w (f f i )/(f f )] (5) WhereS i represents the distance rate of the ith alternative to the positive ideal solution (best cobination), R i represents the distance rate of the ith alternative to the negative ideal solution (worst cobination). The excellence ranking will be based on S i values and the worst rankings will be based on R i values. In other words,s i, R i indicate L 1i and L 0i of L p - etric respectively. 5) Calculate the VIKOR values Q i for i=1,2,,, which are defined as Q i = v [ S i S S S ] + (1 v) [ R i R R R ] (6)

8 WhereS = Max i S i S* = Min i S i, R = Max i R i, R* = Min i R i,, and v is the weight of the strategy of the aority of criteria (or the axiu group utility ). [(S - S*)/(S - S*)] represents the distance rate fro the positive ideal solution of the ith alternative s achieveents In other words, the aority agrees to use the rate of the ith. [(R - R*)/(R - R*)] represents the distance rate fro the negative ideal solution of the ith alternative; this eans the aority disagreewith the rate of the ith alternative. Thus, when the v is larger (>0.5), the index of Q i will tend to aority agreeent; when v is less (<0.5), the index Q i will indicate aority negativeattitude; in general, v = 0.5, i.e. coproise attitude of evaluation experts. 6) Rank the alternatives by Q i values According to the Q i values calculated by step (4), we can rank the alternatives and to ake-decision. 3.Results 3.1.List of final selection: 1. The Rasus: Albus: Peep 1996 Playboys 1997 Hell of a tester 1998 Into 2001 Dead letters 2003 Hide fro the sun 2005 Black roses 2008 Genres: Alternative Rock, Alternative Metal, Funk Rock 2. H.I.M: Albus: 1. Greatest love songs Razorblade roance Deep shadows and brilliant highlights Love etal Dark light Venus doo Screa works: Love in theory and practice 2010 Genres: Gothic Rock, Gothic Metal, Alternative Rock, Hard Rock, Alternative Metal, Heavy Metal 3. Night Wish: Albus: 1. Angels fall first Ocean born Wish aster Century child 2002

9 5. Once Dark passion play 2007 Genres: Syphonic Metal, Power Metal 3.2.List of final criteria Questionnaire three shows iportant criteria for evaluating bands in which fifteen criteria are selected but after the forth questionnaire the nuber of criteria decreased to five that consists of: 1. Technical Ability: Quality of each person in plays their instruent. The ost iportant thing is this criterion is coordinating of group s eber to copose high quality usic.2. Coposing: Quality of coposing. This criterion ephasis on grow of elody in albus of groups. Albus should consider both of rich elody and interest of people and fan too. 3. Moral, Behavior and Lyrics: Faous groups are as a pattern for people and society and its iportant for the to regard to oral, Behavior and one of iportant this for a rock band is to sing lyrics that have huanis contents.4. Ideology and Innovation: It s iportant for any group in the society to have an explicit ai and ideology and proote oral, Behavior and huanitarian in society and the iportant thing in this criterion is innovation that distinct group fro other groups that should be in good way.5. Vocalist: One of iportant thing that can cause a band be faous in the world is vocalist. Voice and ability in singing and acting are very iportant for a vocalist and band. Selecting the best criteria Technical Ability Coposing Moral, Behavior and Lyrics Ideology and Innovation Vocalist TABLE2. Pair wise Matrix of Fuzzy AHP Technical Ability Figure 4: Criteria of selection of best ulti-role artists Coposing Moral, Behavior and Lyrics Ideology and Innovation Vocalist Technical Ability (1,1,1) (2/5,1/2,2/3) (1,3/2,2) (5/2,3,7/2) (2,2/5,3) Coposing (3/2,2,5/2) (1,1,1) (1/5,2,2/5) (1,1/5,2) (1/2,1,3/2) Moral, Behavior (1/2,2/3,1) (5/2,1/2,5) (1,1,1) (2,2/5,3) 1/5,2,2/5) and Lyrics Ideology and Innovation (2/7,1/3,2/5) (1/2,5,1) (1/3,5/2,1/2) (1,1,1) (1,1/5,2) Vocalist (1/3,5/2,1/2) (2/3,1,2) (5/2,1/2,5) (1/2,5,1) (1,1,1) n [ i=1 =1 M gi ] 1 = (0.024, 0.028,0.04) STechnical Ability = (6.9, 6.4, ) (0.024, 0.028, 0.040) = (0.166, 0.179, 0.407) SCoposing= (4.2, 6.2, 7.4) (0.024, 0.028, 0.040) = (0.1, 0.174, 0.296) SMoral,Behavior and Lyrics= (6.2, 4.567, 10.4) (0.024, 0.028, 0.040) = (0.149, 0.128, 416) SIdeology and Innovation= (3.119, 9.033, 4.9) (0.024, 0.028, 0.040) = (0.075, 0.253, 0.196) SVocalist= (5, 10, 9.5) (0.024, 0.028, 0.040) = (0.12, 0.28, 0.38) Min V (STechnical Ability SCoposing, SMoral,Behavior and Lyrics, SIdeology and Innovation,SVocalist) = Min V (SCoposing STechnical Ability, SMoral,Behavior and Lyrics, SIdeology and Innovation, SVocalist) = 0.611

10 Min V (SMoral,Behavior and Lyrics STechnical Ability, SCoposing, SIdeology and Innovation, SVocalist) = Min V (SIdeology and Innovation STechnical Ability, SCoposing, SMoral,Behavior and Lyrics, SVocalist) = Min V (SVocalist STechnical Ability,SCoposing, SMoral,Behavior and Lyrics, SIdeology and Innovation ) = 1 W / = (0.740, 0.611, , 0.738, 1) W = (0.197, 0.163, 0.176, 0.197, 0.267) TABLE 3: Final results of VIKOR Technical Ability Coposing Moral, Behavior and Lyrics Ideology and Innovation Vocalist Weights of F- AHP A Alternatives A A f f Alternatives S i R i V Q i Ranking A A A NOTE: A 1=TheRasus, A 2 = H.I.M, A 3= Night Wish Colclusion: This research used Delphi ethod and MCDM ethods (Fuzzy AHP,VIKOR) together for evaluating and ranking Finnish rock bands in 2000s. At three phases 81 persons participated in this research and they selecting three of the best rock bands with Delphi ethod and after that with two other steps selected five criteria for selected the best bands. Results of Fuzzy AHP ethod showed iportance of criteria and ranked the: 1. Technical Ability 2. Coposing 3.Moral, Behavior and Lyrics 4.Ideology and Innovation 5.Vocalist. Resulted of VIKOR ethod was showed ranking of the Bands: 1. The Rasus2. H.I.M 3. Nightwish. In this evaluation group decision aking was used for calculating of Fuzzy AHP and VIKOR ethods. One of the biggest liitations of this research was selecting and coordinating of experts that took uch tie. This study can be used as a pattern and fraework for awards. Certainly for using this study, should pay ore attention for selecting the nuber of experts. Because of liitation in this research used only 81 persons. And at the end this study can be used as a pattern for selecting other groups of artists and another countries, ale singers, feale singers and etc. References [1]. Bacchiocchi, S The Worldview Of Rock Music, Endtie Issues No. 33, 15 Deceber 1999.

11 [2]. Beebe, R. et all, Rock Over the Edge: Transforations in Popular Music Culture (Durha, NC: Duke University Press, 2002), ISBN X, p. 7. [3]. Binkley J, Finch E, Hall J, Black T, Gowland C. Diagnostic classification of patients with low back pain: report on a survey of physical therapy experts. PhysTher 1993;73: [4]. Brake, M Coparative Youth Culture: the Sociology of Youth Cultures and Youth Subcultures in Aerica, Britain, and Canada (Abingdon: Routledge, 1990), ISBN , pp and [5]. Buckley, J. J. (1985). Fuzzy hierarchical analysis. Fuzzy Sets and Systes, 17(3), [6]. Chang, D. Y. (1996). Applications of the extent analysis ethod on fuzzy AHP.European Journal of Operational Research, 95(3), [7]. Cheng, C. H. (1997). Evaluating naval tactical issile systes by fuzzy AHP based on he grade value of ebership function. European Journal of OperationalResearch, 96(2), [8]. Dag deviren, M., Yüksel, _I.,& Kurt, M. (in press). A fuzzy analytic network process (ANP) odel to identify faulty behavior risk (FBR) in work syste.safetyscience. [9]. Dalkey, N. C., &heler, O. (1963). An experiental application ethod to the use of experts. Manageent Science, 9(3), [10]. Dalkey, N. Heler, O An experiental application of the Delphi ethod to the use of experts, Manageent Science 9 (3), 1963, pp [11]. Delbecq, A.L. et all, Group Techniques for Progra Planning: A Guide to Noinal Group and Delphi Processes, Scott, Foresan and Copany, Glenview, Illinois, [12]. Deng, H. (1999). Multicriteria analysis with fuzzy pairwise coparison. International Journal of Approxiate Reasoning, 21(3), [13]. Hashekhani Zolfani, S. et all, Using hybrid MCDM approach for selecting the best ultirole artist of rock bands in 2000s, International Journal of Manageent and Innovation, Volue 3, Issue 2, [14]. Hayne, S. Pollard, C A coparative analysis of critical issues facing Canadian inforation systes personnel: a national and global perspective, Inforation & Manageent 38 (2), 2000, pp [15]. Holsapple, P. Joshi, K Knowledge anipulation activities: results of a Delphi study, Inforation & Manageent 39 (6), 2002, pp [16]. Jang, D. et all, Music Genre Classification Using Novel Features and A Weighted Voting Method, Proceedings of the ICME 2008, Hannover, Gerany. [17]. Lee, A. H. I., Chen, W.-C., & Chang, C.-J. (2008). A fuzzy AHP and BSC approach for evaluating perforance of IT departent in the anufacturing industry intaiwan. Expert Systes with Applications, 34(1), [18]. Leung, L. C., & Cao, D. (2000). On consistency and ranking of alternatives in fuzzy AHP. European Journal of Operational Research, 124(1), [19]. Liang, G. S., & Wang, M. J. (1994). Personnel selection using fuzzy MCDM algorith. European Journal of Operational Research, 78,

12 [20]. Mikhailov, L. (2004). A fuzzy approach to deriving priorities fro interval pairwisecoparisonudgents. European Journal of Operational Research, 159(3), [21]. Murry, J. W., & Haons, J. O. (1995). Delphi: A versatile ethodology for conducting qualitative research. The Review of Higher Education, 18(4), [22]. Mulligan, P Specification of a capability-based IT classification fraework, Inforation & Manageent 39 (8), 2002, pp [23]. Opricovic, S. Multi-criteria optiization of civil engineering systes, Faculty of Civil Engineering, Belgrade, [24]. Opricovic, S., &Tzeng, G. H, Multicriteria planning of post earthquake sustainable reconstruction, Coputer-Aided Civil and Infrastructure Engineering, no.17, pp , [25]. Okoli, C. Pawlowski, S.D The Delphi ethod as a research tool: an exaple, design considerations and applications, Inforation & Manageent 42 (2004) [26]. PeMyint, E.E and Pwint, M An Approach for Mulit-Label Music Mood Classification nd International Conference on Signal Processing Systes (ICSPS), C 2010 IEEE. [27]. Rowe, G. et all, Delphi: a re-evaluation of research and theory, Technological Forecasting and Social Change, 1991, pp [28]. Schidt, R.C Managing Delphi surveys using nonparaetric statistical techniques, Decision Sciences 28 (3), 1997, pp [29]. Van Laarhoven, P. J. M., &Pedrycz, W. (1983). A fuzzy extension of Saaty s priority theory. Fuzzy Sets and Systes, 11(1 3), [30]. Yu, C. S. (2002). A GP-AHP ethod for solving group decision-aking fuzzy AHPprobles. Coputers and Operations Research, 29,