A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss JOURNAL OF AERONAUTICS AND SPACE TECHNOLOGIES JANUARY 2014 VOLUME 7 NUMBER 1 (45-52) A MULTI-CRITERIA CALL CENTER SITE SELECTION BY HIERARCHY GREY RELATIONAL ANALYSIS Semra BİRGÜN Cegz GÜNGÖR * Beyket Uversty, Istabul, Turkey semrabrgu@beyket.edu.tr Aeroautcs ad Space Techologes Isttute, Turksh Ar Force Academy, Istabul, Turkey ggrcgz@gmal.com Receved:,31 th July 2013 Accepted: 30 th Jauary 2014 ABSTRACT Faclty locato selecto s a decso problem that takes to cosderato both qualtatve ad quattatve factors ad also s oe of the most mportat strategc decsos affectg orgazatos terms of busess success. Fudametally, t comprses evaluatg a group of alteratve stes o the bass of multple crtera. I ths cotext whe alteratves ad factors are multtudous, mult-crtera decso makg methods are beg used for successful decsos. I ths paper the decso problem s exemplfed by applyg o a part of a proect for a call ceter ste selecto by usg oe of the mult-crtera decso-makg methods, amely herarchy grey relatoal aalyss, based o applcato of aalytc herarchy process (AHP) ad grey relatoal aalyss (GRA) methods. The goal of the selecto s to determe the most approprate locato amog the alteratves. Ne ctes gve as alteratves are evaluated ad compared agast huma resources, ecoomc ad regoal codtos cludg fourtee sub-crtera. Keywords: Aalytc Herarchy Process, Grey Relatoal Aalyss, Call Ceter Ste Selecto. HİYERARŞİK GRİ İLİŞKİSEL ANALİZ YÖNTEMİYLE ÇOK KRİTERLİ ÇAĞRI MERKEZİ YERİ SEÇİMİ ÖZET Tess kuruluş yer seçm tel ve cel faktörler dkkate ala br karar problem olmakla beraber ayı zamada orgazasyou başarısıı etklemes açısıda da e öeml stratek kararlarda brsdr. Esasıda br grup alteratf brçok krter temelde değerledrmey çermektedr. Bu kapsamda alteratf ve etkeler fazla olduğu zama başarılı kararlar ç çok krterl karar verme yötemler kullaılmaktadır. Bu makalede karar problem, Aaltk Hyerarş Proses (AHP) ve Gr İlşksel Aalz (GİA) yötemler temel ala çok krterl karar verme yötemlerde Hyerarşk Gr İlşksel Aalz yötem çağrı merkez yer seçm ç br proe parçası olarak uygulamasıyla örekledrlmştr. Seçm amacı alteratfler arasıda e uygu yer belrlemektr. Alteratf olarak verle dokuz şehr odört alt krter çere sa kayağı, ekoomk ve bölgesel şartlar kapsamıda karşılaştırılmış ve değerledrlmştr. Aahtar Kelmeler: Aaltk Hyerarş Proses, Gr İlşksel Aalz, Çağrı Merkez Kuruluş Yer Seçm. 1. INTRODUCTION Decso makg s oe of the most mportat problem ay feld ad t s the process of fdg the best opto from all of the feasble alteratves. Most decso makg problems ofte have multple ad cotradctory evaluato stadards. Varous opos amog decso makers are the ma cause cotrbutg to the coflcts the process of decso makg. I real world stuatos, because of defcet * Correspodg Author 45 or o obtaable formato, the attrbutes are ofte ot so determstc. But the maorty of these attrbutes ca be assessed by huma percepto ad huma udgmet. Faclty locato selecto s a decso problem that takes to cosderato both qualtatve ad quattatve factors ths respect ad s oe of the most mportat strategc decso affectg orgazatos substatally, compettveess ad
A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss performace. It would be ot oly a challegg decso for the compaes but also too costly ad too dffcult to chage the ste after the stallato of a plat. Nowadays umerous compaes are beeftg from umercal decso makg methods the decso makg process of ste selecto because ths strategc decso allow them to perform operatos wth mmum cost ad maxmum proft ad also to mata ther presece effectvely. Icludg dverse factors, faclty locato selecto desco s a process that requres selecto of the ratoal processes amog alteratves. I ths cotext whe alteratves ad factors are multtudous, mult-crtera decso makg methods are beg used for successful decsos. I ths study faclty locato selecto problem s hadled ad solved for the call ceters, cosdered to ope the ear future southeaster Aatola, usg a combato of the Aalytc Herarchy Process (AHP) ad the Grey Relatoal Aalyss (GRA). The selecto of call ceter locato s a complex mult-crtera task whch cludes both quattatve ad qualtatve factors are coflct ad ucerta. For the purpose of solvg the problem all crtera whch affect the decso makg process are determed by a group of experts occuped o call ceter sector as executves. The most approprate locato amog the alteratves s set the decso makg process ad created by examg call ceter faclty locato selecto crtera. 2. LITERATURE REVIEW Selecto problem has bee terested by may researchers, thus umerous optmzato models have bee developed the varous studes for the ste selecto problems durg the past years. Ths paper deals wth a approach based o AHP ad GRA for choosg the best call cetre ste. Aalytc Herarchcal Process (AHP) whch s a specal case of the ANP, has bee wdely used for locato problems, cludg Aras et al. [1], whch a pretty umber of crtera were take to accout for a wd observato stato locato problem. Aother example of AHP locato problems s Tzeg et al. [2] whch 4 alteratves, 5 aspects ad 11 crtera were used for a locato evaluato of a restaurat. I ths paper, the compromse rakg method, amed VIKOR, has bee troduced as oe applcable techque. The VIKOR algorthm determes the weght stablty tervals, for the obtaed compromse soluto wth the put weghts, dcatg the preferece stablty of obtaed compromse soluto. Feradez ad Ruz [3] cosdered the selecto of a locato for a dustral park. I ther paper, they have proposed a three-level herarchcal decso process whch each level has ts ow geographcal decso crtera. They the used AHP to fd the locato. Goal programmg has bee utlzed to mprove the problems solved by AHP. For example, Badr [4] offered a combed AHP ad goal program modelg approach for teratoal faclty locato/allocato problem; the role of AHP was to prortze the set of locato alteratves at frst. Uder may stuatos, the values of the quattatve ad qualtatve crtera are ofte mprecse or vague, therefore GRA, oe of the sub-braches of Deg s Grey Theory [5] whch has bee appled predcto, cotrol, socal ad ecoomc system maagemet, decso makg about evrometal systems recet years [6-9], s becomg a hady approach, lke Zhe-qag et al. [10] whch preseted a aalyss for the faclty s locato of logstcs dstrbuto etwork. Huag ad Huag [11] tegrated fuzzy ad grey modelg methods for predctg the mothly average temperatures Tape. The basc grey model GM(1,1) s accompaed wth the adaptve fuzzy method to mprove ts predcto capablty. They foud that the predctve capablty of the tegrated model was satsfactory for those systems demadg complcated cotrol varables ad rules. I aother study Chag ad L [12] used GRA to aalyze how eergyduced CO 2 emssos from 34 dustres Tawa are affected by the factors: producto, total eergy cosumpto, coal, ol, gas ad electrcty uses. Results of the study dcated that dustral producto has the closest relatoshp wth aggregate CO 2 emsso chages; electrcty cosumpto the secod mportace. They poted out the ecoomy Tawa reled heavly o CO 2 tesve dustres, ad that electrcty cosumpto had become more mportat for ecoomc growth. Aother example s Kahrama et al. [13] whch to select techology for reewable electrcty geerato, mplemeted AHP ad GRA for ther mult-crtera decso makg. They obtaed best alteratve by evaluatg the problem uder 3 crtera, 10 sub crtera. The result showed that photovoltac power s the optmal alteratve for vestg the dfferet reewable electrcty geerato techologes. Yag ad Che [14] used a combed AHP ad GRA for suppler selecto problem. They used AHP to calculate relatve mportace weghtgs of qualtatve crtera. The, the qualtatve ad quattatve data were utlzed together ad obtaed the grey relatoal grade values. The best suppler had the hghest grey relatoal value amog others. Zeg et al. [15] employed a approach for the waste water treatmet alteratve selecto problem. Ths was based o AHP ad GRA. They used 3 ma attrbutes cludg 8 dces that represeted the alteratves ad evaluated four water treatmet methods. Feg et al. [16] preseted a study based o establshg a evaluato dex system of logstcs ceter locato. For ths purpose they costructed a tegrated decso model 46
A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss by usg the etropy method ad grey relatoal aalyss. The weghts of the evaluato dexes were defed by the etropy method. The quattatve process ad comparso of the qualtatve formato were made by GRA. 3. METHODOLOGY: INTEGRATED AHP AND GRA A. AHP Procedure Aalytc herarchy process (AHP) was troduced by Saaty [17] ad afterwards t gaed wdely acceptace [1-4], [13-15]. AHP has bee used to solve multple crtera decso makg problems dfferet areas of huma eeds ad terests. The herarchy s costructed such a way that the overall decso goal s at the top level, decso crtera are the mddle level(s), ad decso alteratves at the bottom [18]. Three steps AHP, decomposto, udgmet ad sytheszg are the same way as people thk. So t could be sad that the AHP s a subectve weghtg method. The relatve mportace betwee two comparatve factors s reflected by the elemet values of udgmet matrx. Table 1 shows geeral form of the measuremet scale. It has relatve mportace scale of 1-9 [17], [19]. Table 1. Scale for parwse comparso AHP. Importace Descrptos degree 1 Equally mportat 3 Weakly mportat 5 Strogly mportat 7 Very strogly mportat 9 Extremely mportat 2, 4, 6, 8 Itermedate values Explaato Crtera ad are of equal mportace Crtera s weakly more mportat tha obectve Crtera s strogly more mportat tha obectve Crtera s very strogly more mportat tha obectve Crtera s extremely more mportat tha obectve For example, a value of 8 meas that Crtera s mdway betwee strogly ad more mportat tha obectve After defg ad decomposg the problem to a herarchcal structure wth decso elemets, the procedures of AHP to solve the problem geerally volve three essetal steps order [20]: 1) The parwse comparso matrx (A) s formed a11 a12 a1 a21 a22 a2 A( a ) x a a a 1 2 (1) where a represets the udgmet degree of th factor compared to th factor. 2) The cosstecy of comparso matrx s computed as follows: maks CI 1 (2) where egevalue close to s the largest egevalue (λ max ) ad ca be foud by eg() structo va Matlab ad CI s the cosstecy dex. Cosstecy check s appled by computg the cosstecy rato (CR): CI CR (3) RI where RI s the radom dex. The values of RI are show Table 2. Table 2. RI values. m 2 3 4 5 6 7 8 R.I. 0 0,58 0,9 1,12 1,24 1,32 1,41 Whe CR 0.10, t meas that the cosstecy of the parwse comparso matrx s the desred terval ad matrx s acceptable. 3) The weghts vector (W A ) s the estmated by usg the egevalue method through the followg formula: W A 1 1 1/ 1/ a11 a12 a1 a2 1 a21 a22 a2 1 a a 1/ (... ) (... ) ( a. a.. a ) 1 2 (3) The ormalzed weghts vector (W ' A ) s the obtaed as follows: W ' A 1/ 1/ a / 1 a 1 1 1 1/ 1/ a / 2 a 1 1 1 1/ 1/ a / a 1 1 1 W1 W2 W (4) 47
A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss B. Grey Relatoal Aalyss The grey relatoal aalyss (GRA) s used to determe the relatoshp (smlarty) betwee two seres of data a grey system. Its structure has ucertaty, therefore t hadles the problems cossted of dscrete data ad partal formato [5]. It operates the grey relatoal grade to determe the relatoal degree of factors. The algorthm of GRA s llustrated as follows [13-16], [20]: 1) Let x 0 deote the referetal seres wth ettes ad let x represet the compared seres. X 0 ( X0(1), X0(2),..., X0( )) (5) X ( X(1), X( ),..., X( )), 1,2,..., m 1,2,..., (6) 2) Before calculatg the grey relatoal grade, we must perform data pre-processg. Normalzato of seres must be doe to esure that all of them are the same order. Normalzed sequeces ca be deoted as: X ( ) ( X (1), X ( ),..., X ( )) * * * * (7) For cost x dces, the ormalzed data ca be acqured by X X ( ) m * m 1, 2,..., 1, 2,..., (8) X ( ) Whle for beeft x dces, the ormalzed data ca be acqured by * X() X ( ) 1,2,..., m 1,2,..., X (9) max 3) For th factor, the grey relatoal coeffcet betwee seres x 0 ad x s the gve as: mmax () (10) 0 () max 0 where, * 0 ( ) X0( ) X ( ), max max max 0 ( ), m m m ( ) ad ρ s the dstgushg 0 coeffcet, ρ Є [0,1], ad typcally ρ = 0,5. 4) Fally, by usg the weghts the aggregated evaluato model ca be wrtte as follows: w ( ) 0 1 4. PRACTICAL CASE (11) I ths secto, the herarchy GRA s appled to the ste selecto of a call ceter whch s gog to be establshed the southeaster Aatola rego accordace wth the opo of the proect executves of a corporato, located Istabul, terested vestg the rego. I ths cotext, e ctes are take to accout ad coded for the smplcty from A1 to A9 alphabetc order. The decso model for the call ceter ste selecto problem s gve Fg. 1. It cotas four levels: at the top of the herarchy, the overall obectve s to select the most approprate ste for the call ceter. The crtera level s the secod level of the herarchy ad cossted of huma resources, ecoomc ad regoal codtos crtero (C1, C2, C3). The thrd level cosdered as dex level cotas dces: populato, o-farm payrolls, educated populato, populato growth rate, youthful populato, uemploymet rate, presece of hgher educato sttutos, umber of employees the sector, come per capta, vestmet cetves, lad cost, labor cost, trasportato, clmate (I 1 to I 14 ). Fally, alteratve level of the model pots out the ctes to be compared ad evaluated. Level 1 Level 2 Level 3 Level 4 ak1 ak2 ak3 Huma ak4 Resources ak5 ak6 A1 ak7 A2 ak8 A3 A4 Selecto ak9 of the Ecoomc A5 best ste ak10 ak11 A6 ak12 A7 A8 Regoal Codtos ak13 ak14 Fgure 1. A herarchy decso model for call ceter ste selecto. Sce the herarchy has bee establshed for the problem, we eed to compute the weghts descrbg the decso makers relatve mportace of ther udgmets o alteratves. Table 3 ad Table 4 dsplays obtaed weghts ad cosstecy ratos ad the values of crtera hadled by decso makers A9 48
A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss respectvely. The results Table 3 llustrate that the weght of the ecoomc crtero s 0.550 as compared to 0.368 for the huma resources crtero, dcatg that the mportace of ecoomc crtero s early more oe ad a half tmes tha huma resources crtero. Ad also as compared to 0.082 for the regoal codtos crtero, ecoomc crtero has early seve tmes more mportace tha regoal codtos crtero. I Table 4 some dces are provded by the umercal values ad some are by the quatfcato of the lgustc values. I some cases ucerta dces, such as clmate, trasportato etc., ca be quatfed. Decso makers ca classfy dces to fve grades wth descrptve laguage cludg excellet, good, moderate, poor ad very poor. Table 3. Crtera ad dce weghts oted by decso makers. Crtera Weght CR Idces Weght CR Populato (I 1 ) 0.390 No-farm payrolls (I 2 ) 0.157 Educated populato (I 3 ) 0.058 Huma Populato growth rate (I Resources 0.368 4 ) 0.131 Youthful populato (I (C1) 5 ) 0.104 Uemploymet rate (I 6 ) 0.082 0.067 Presece of hgher educato sttutos (I 7 ) 0.026 0.074 Number of employees the sector (I 8 ) 0.052 Icome per capta (I 9 ) 0.571 Ecoomc (C2) Regoal Codtos (C3) 0.550 0.082 Ivestmet cetves (I 10 ) 0.044 Lad cost (I 11 ) 0.253 Labor cost (I 12 ) 0.132 Trasportato (I 13 ) 0.833 Clmate (I 14 ) 0.167 Table 4. A obectve herarchy for call ceter ste selecto. Alteratves Crtera Idces A1 A2 A3 A4 A5 A6 A7 A8 A9 I 1 595261 534205 1592167 1799558 124320 773026 1762075 466982 310879 I 2 118856 76564 198272 340341 24803 113874 199728 63374 44189 I 3 103028 77633 232994 262940 7393 102154 158493 53595 33077 C1 I 4 P(0.3) G(0.7) M(0.5) G(0.7) P(0.3) M(0.5) G(0.7) G(0.7) P(0.3) I 5 117285 112399 333345 316305 22944 164997 350141 111396 70153 I 6 10.0 11.4 13.2 13.1 9.9 8.9 12.1 10.9 12.4 I 7 1 1 2 3 1 1 1 1 1 I 8 E(0.9) E(0.9) M(0.5) E(0.9) E(0.9) E(0.9) E(0.9) E(0.9) E(0.9) I 9 9521 10609 10678 11022 11397 9164 8041 6068 9115 C2 I 10 G(0.7) E(0.9) E(0.9) P(0.3) G(0.7) E(0.9) E(0.9) E(0.9) E(0.9) I 11 110 265 250 149 163 368 168 107 60 I 12 G(0.7) G(0.7) G(0.7) G(0.7) G(0.7) G(0.7) G(0.7) G(0.7) G(0.7) C3 I 13 M(0.5) M(0.5) E(0.9) G(0.7) M(0.5) G(0.7) G(0.7) P(0.3) M(0.5) I 14 G(0.7) G(0.7) G(0.7) E(0.9) E(0.9) E(0.9) E(0.9) P(0.3) M(0.5) 0.067 0 Accordgly, the subecto grade s 0.9, 0.7, 0.5, 0.3 ad 0.1, respectvely [21]. The data Table 4 was studed for the purpose of applyg herarchy GRA. Eq. 8 s used for the cost dces (I 8, I 9, I 11, I 12 ) ad Eq. 9 for the rest of the dces as beeft formula. The ormalzed values of all dces ca be foud Table 5. Table 5 shows the requred data for computato of prmary ad secodary grey relatoal coeffcets. These are calculated by usg Eq. 10 ad ρ as 0,5. Acheved data are dsplayed Table 6 ad Table 7. At the ed, the aggregated grey relatoal grade vector ca be obtaed by multplyg the resultg secodary grey relatoal coeffcet matrx Table 7 by the weghtg vector as show Eq. 11 for the crtero level (level 2) wth respect to the overall obectve. As llustrated Table 8, the e alteratve stes, that s A1 (Adıyama), A2 (Batma), A3 (Dyarbakır), A4 (Gazatep), A5 (Kls), A6 (Mard), A7 (Şalıurfa), A8 (Şırak) ad A9 (Srt), are raked 7, 8, 3, 1, 9, 6, 2, 4 ad 5, respectvely. Therefore, Gazatep as A4 s the optmal alteratve amog the others for the call ceter ste. 49
A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss Table 5. Normalzed data of alteratves for dex level. Alteratves Crtera Idces A1 A2 A3 A4 A5 A6 A7 A8 A9 I 1 0.33 0.30 0.88 1.00 0.07 0.43 0.98 0.26 0.17 I 2 0.35 0.22 0.58 1.00 0.07 0.33 0.59 0.19 0.13 I 3 0.39 0.29 0.87 0.98 0.03 0.38 0.59 0.20 0.12 C1 I 4 0.33 0.78 0.56 0.78 0.33 0.56 0.78 0.78 0.33 I 5 0.33 0.32 0.95 0.90 0.07 0.47 1.00 0.32 0.20 I 6 0.59 0.67 0.78 0.77 0.58 0.52 0.71 0.64 0.73 I 7 0.33 0.33 0.67 1.00 0.33 0.33 0.33 0.33 0.33 I 8 0.11 0.11 0.20 0.11 0.11 0.11 0.11 0.11 0.11 I 9 0.53 0.48 0.47 0.46 0.44 0.55 0.63 0.83 0.56 C2 11 I 10 0.78 1.00 1.00 0.33 0.78 1.00 1.00 1.00 1.00 I 0.51 0.21 0.22 0.38 0.34 0.15 0.33 0.52 0.93 I 12 0.43 0.43 0.43 0.43 0.43 0.43 0.43 0.43 0.43 C3 I 13 0.56 0.56 1.00 0.78 0.56 0.78 0.78 0.33 0.56 I 14 0.78 0.78 0.78 1.00 1.00 1.00 1.00 0.33 0.56 Table 6. Prmary grey relatoal coeffcets for dex level. Alteratves Crtera Idces A1 A2 A3 A4 A5 A6 A7 A8 A9 I 1 0.42 0.41 0.81 1.00 0.34 0.46 0.96 0.40 0.37 I 2 0.43 0.39 0.54 1.00 0.35 0.42 0.54 0.38 0.36 I 3 0.44 0.41 0.79 0.97 0.33 0.44 0.55 0.38 0.36 C1 I 4 0.42 0.69 0.52 0.69 0.42 0.52 0.69 0.69 0.42 I 5 0.42 0.42 0.91 0.84 0.34 0.48 1.00 0.42 0.38 I 6 0.54 0.60 0.69 0.68 0.54 0.51 0.63 0.58 0.64 I 7 0.42 0.42 0.59 1.00 0.42 0.42 0.42 0.42 0.42 I 8 0.36 0.36 0.38 0.36 0.36 0.36 0.36 0.36 0.36 I 9 0.51 0.48 0.48 0.47 0.47 0.52 0.57 0.75 0.52 C2 11 I 10 0.69 1.00 1.00 0.42 0.69 1.00 1.00 1.00 1.00 I 0.50 0.38 0.39 0.44 0.43 0.37 0.42 0.51 0.87 I 12 0.46 0.46 0.46 0.46 0.46 0.46 0.46 0.46 0.46 C3 I 13 0.52 0.52 1.00 0.69 0.52 0.69 0.69 0.42 0.52 I 14 0.69 0.69 0.69 1.00 1.00 1.00 1.00 0.42 0.52 Table 7. Secodary grey relatoal coeffcets for crtero level. Weghted prmary grey relatoal coeffcets A1 A2 A3 A4 A5 A6 A7 A8 A9 C 1 0.429 0.458 0.701 0.881 0.371 0.460 0.767 0.449 0.398 C 2 0.509 0.475 0.477 0.459 0.468 0.495 0.536 0.662 0.622 C 3 0.548 0.548 0.948 0.742 0.600 0.742 0.742 0.420 0.520 Secodary grey relatoal coeffcets A1 A2 A3 A4 A5 A6 A7 A8 A9 C 1 0.370 0.385 0.595 1.000 0.342 0.387 0.698 0.381 0.355 C 2 0.466 0.437 0.439 0.424 0.431 0.454 0.494 0.671 0.602 C 3 0.417 0.417 1.000 0.581 0.451 0.581 0.581 0.351 0.400 Table 8. Grey relatoal grades for alteratves. Alteratves Grey relatoal grade Rak A1 0.427 7 A2 0.416 8 A3 0.542 3 A4 0.649 1 A5 0.400 9 A6 0.440 6 A7 0.576 2 A8 0.538 4 A9 0.495 5 50
A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss 5. CONCLUSION The call ceter ste selecto problem s a dffcult mult-crtera decso makg process to hadle. The most crucal features of ths process are complexty ad ucertaty. As a ovel approach for soluto, herarchy GRA (HGRA), based o AHP ad GRA, has bee utlzed to determe the faclty locato for call ceter. The proposed model comprses two parts. The frst part apples covetoal AHP to determe the relatve weghts of the crtera. Ad the secod part apples GRA to rak the alteratves ad the selects the optmum ste for call ceter. The dfferet prortes gve to the crtera by experts or decso makers are reflected through the weghts, so the bas arsg from subectve udgmets ad radom effects ca be preveted. The ctes southeaster Aatola are used to demostrate the effectveess of the proposed methodology for selectg the best call ceter ste. The method provdes a obectve ad effectve decso model for selectg the most approprate locato. The aalytcal results reveal that such a approach ca cope wth complcated mult-crtera decso makg processes ad provde scetfc ad reasoable results for decso makers. Furthermore, compaes, local polcy makers ad other decso makers ca use ths method ay feld relato to mult-crtera decso makg problems. 6. REFERENCES [1] Sa Crstóbal, J. R., (2011) Mult-crtera Decso-makg the Selecto of a Reewable Eergy Proect Spa: The VIKOR Method, Reewable Eergy, 36(2): 498-502. [2] Tzeg, G. H., Teg, M. H., Che, J. J., Oprcovc, S., (2002) Multcrtera Selecto for a Restaurat Locato Tape, Iteratoal Joural of Hosptalty Maagemet, 21(2): 171-187. [3] Feradez, I., Ruz M. C., (2009) Descrptve Model ad Evaluato System to Locate Sustaable Idustral Areas, Joural of Cleaer Producto, 17(1): 87-100. [4] Badr, M. A., (1999) Combg the Aalytc Herarchy Process ad Goal Programmg for Global Faclty Locato-Allocato Problem, It. J. Producto Ecoomcs, Vol. 62: 237-248. [5] Deg, J. L., (1989) Itroducto to Grey System, Joural of Grey System, 1(1), 1-24. [6] We, K. L., (2004) The Grey System Aalyss ad Its Applcato Gas Breakdow ad VAR Compesator Fdg (vted paper), Iteratoal Joural of Computatoal Cogto, 2(1): 21 44. [7] Che, F. L., Che, Y. C., (2010) Evaluatg the Mateace Performace of the Semcoductor Factores Based o the Aalytcal Herarchy Process ad Grey Relatoal Aalyss, Amerca Joural of Appled Sceces, 7(4): 568-574. [8] Karmakar, S., Muumdar, P. P., (2006) Grey Fuzzy Optmzato Model for Water Qualty Maagemet of a Rver System, Advaces Water Resources, 29(7): 1088-1105. [9] Zeg, J., Jag, Z., L, T., Yag, D., Km, Y. H., (2011) Applcato of Etropy Weght Techque Grey Relatoal Aalyss for Scheme Optmzato of Sewage Treatmet Plat, Advaced Materals Research, Vol. 199-200: 1722-1728. [10] Zhe-qag, B., Peg, W., Fag, Y., Cog-we, Z., Le, G., (2008) Grey Relato Degree Aalyss for the Faclty's Locato of Logstcs Dstrbuto Network, Iteratoal Symposums o Iformato Processg, pp. 615-619. [11] Huag, Y. P., Huag, C. C., (1996) The Itegrato ad Applcato of Fuzzy ad Grey Modelg Methods, Fuzzy Sets ad Systems, 78(1): 107-119. [12] Chag, T. C., L, S. J., (1999) Grey Relato Aalyss of Carbo Doxde Emssos from Idustral Producto ad Eergy Uses Tawa, Joural of Evrometal Maagemet, 56(4): 247-257. [13] Saruca, A., Baysal, M. E., Kahrama, C., Eg, O., (2011) A Herarchy Grey Relatoal Aalyss for Selectg the Reewable Electrcty Geerato Techologes, Proceedgs of the World Cogress o Egeerg, Vol. 2: 1149-1154. [14] Yag, C. C., Che, B. S., (2006) Suppler Selecto Usg Combed Aalytcal Herarchy Process ad Grey Relatoal Aalyss, Joural of Maufacturg Techology Maagemet, 17(7): 926 941. [15] Zeg, G., Jaga, R., Huaga, G., Xua, M., La, J., (2007) Optmzato of Wastewater Treatmet Alteratve Selecto by Herarchy Grey Relatoal Aalyss, Joural of Evrometal Maagemet, Vol. 82, 250-259. [16] Feg, L., Yua, R., (2011) Study o Grey Relato Aalyss Based o Etropy Method Evaluato of Logstcs Ceter Locato, Thrd Iteratoal Coferece o Measurg Techology ad Mechatrocs Automato, Vol. 3: 474-477. [17] Saaty, T. L., (1980) The Aalytc Herarchy Process, McGraw-Hll, New York. [18] Wag, Y. M., Lu, J., Elhag, T. M. S., (2008) A Itegrated AHP - DEA Methodology for Brdge Rsk Assessmet, Computers & Idustral Egeerg, 54(3): 513-525. [19] Xu, G., Yag, Y. P., Lu, S. Y., L, L., Sog, X., 51
A Multcrtera Call Ceter Ste Selecto by Herarchy Grey Relatoal Aalyss Comprehesve Evaluato of Coal-Fred Power Plats Based o Grey Relatoal Aalyss ad Aalytc Herarchy Process, Eergy Polcy, 39(5): 2343-2351. [20] Gügör, C., (2013) A Mult Crtera Faclty Locato Selecto By Herarchy Grey Relatoal Aalyss Method, (master thess), Turksh Ar Force Academy Aeroautcs ad Space Techologes Isttute, İstabul, Turkey. [21] Kaya, T., Kahrama, C., (2010) Multcrtera Reewable Eergy Plag Usg A Itegrated Fuzzy VIKOR&AHP Methodology: The Case of Istabul, Eergy, Vol. 35: 2517-2527 VITAE Prof.Dr. Semra BİRGÜN She graduated from Yıldız Techcal Uversty (YTU) 1980. She receved her M.S. degree 1983 ad her PhD degree 1989 from Maagemet Egeerg Departmet at Istabul Techcal Uversty (ITU). She became assocate professor brach of Idustral Egeerg 1993 ad professor 2002. She studed the Departmet of Idustral Egeerg at YTU 1982-1989, the Departmet of Maagemet Egeerg at ITU 1989-2002, the Departmet of Idustral Egeerg at Kocael Uversty 2002-2004 ad the Departmet of Idustral Egeerg at Istabul Commerce Uversty, respectvely. She has bee charge of at Beyket Uversty as head of the Maagemet Egeerg Departmet sce 2011. Maufacturg Systems, Producto Plag ad Cotrol, Lea Maufacturg, Group Techology ad Cellular Maufacturg, Value Stream Mappg ad Theory of Costrats are some of the felds she s terested. She has a umber of atoal ad teratoal publcatos. 1 st Lt. Cegz GÜNGÖR He was bor Mers, Turkey. He obtaed hs B.S degree from Idustral Egeerg Departmet at Turksh Ar Force Academy (TUAFA) 2006, M.S degree from Aeroautcs ad Space Techologes Isttute (ASTIN) 2013. Hs research terests are operatos research ad mult crtera decso makg methods. 52