Food Sci Nutr 2017; 1–11  | 1www.foodscience-nutrition.com

Received:18October2016  |  Revised:22November2016  |  Accepted:24November2016DOI:10.1002/fsn3.454

O R I G I N A L R E S E A R C H

Sensory evaluation of selected formulated milk barberry drinks using the fuzzy approach

Zahra Tahsiri | Mehrdad Niakousari | Sara Khoshnoudi-Nia | Seyed Mohamad H. Hosseini

ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium,providedtheoriginalworkisproperlycited.©2017TheAuthors.Food Science & NutritionpublishedbyWileyPeriodicals,Inc.

DepartmentofFoodScienceandTechnology,SchoolofAgriculture,ShirazUniversity,Shiraz,Iran

CorrespondenceMehrdadNiakousari,DepartmentofFoodScienceandTechnology,SchoolofAgriculture,ShirazUniversity,Shiraz,Iran.Email:niakosar@shirazu.ac.ir,mehrnia2012@yahoo.com

AbstractAmidrigidcompetitioninmarketingtoaccomplishcustomers’needs,thecostofdis-appointment is toohigh. In aneffort toescapemarketdisappointment,oneof theoptionstobeconsideredisprobingforcustomersatisfactionthroughsensoryevalua-tion.Thisstudyaimstorankthesixselectedmilk-barberrydrinkformulaeoutof24(codenumbersS3,S4,S15,S16,S17andS18)eachhavingdifferentmilk:barberry:pectinamount(7:3:0.2;6:4:0.2;7:3:0.4,6:4:0.4,5:5:0.4and6:4:0.4),respectively,andtodeterminethebestofqualityattributethroughsensoryevaluation,usingthefuzzydecision-makingmodel.TheselectionwasbasedonpH,totalsolidcontent,andde-greeofserumseparationandrheologicalpropertiesofthedrinks.Theresultsshowedthat the S4 had the highest acceptability, rated under the “very good” category,whereasthelowestacceptabilitywasreportedfortheS3whichwasclassifiedunderthe“satisfactory”category.Insummary,therankingofthemilk-barberrydrinkswasS4>S17>S16>S15>S18>S3.Furthermore,qualityattributeswererankedastaste>mouthfeel>aroma>color.Resultssuggestthatthefuzzyapproachcouldbeappropriatelyusedtoevaluatethistypeofsensorydata.

K E Y W O R D S

functionaldrinks,fuzzydecisionmaking,milk-barberrydrinks,pectin,sensoryevaluation

1  | INTRODUCTION

Antioxidants principally function to reduce oxidizing damages, asthese damages contribute to the cause of cardiovascular disease,Alzheimer’s,cancer,cataractanddiabetes(Rodríguez-Roque,Rojas-Graü,Elez-Martínez,&Martín-Belloso,2013;Slatteryetal.,2000).Fruits andvegetables contain different bioactive compounds suchasvitaminsA,C,andE.Phenoliccompoundsarealsofoundinfruitswith antioxidant activities, and have shown to be good contribu-torstothetotalantioxidantcapacityofthefoodthatcontainthem(Chaovanalikit &Wrolstad, 2004). Barberry is rich in anthocyaninand vitamin C. Anthocyanins are polyphenol compounds whichbelongtothegroupofwatersolublepigmentsthatcanbeusedfor

food coloring (Harborne, 2013). Recently, considerable attentionhas beendirected to nutraceutical foodswhichhave ledbreedersto initiatetheselectionofplantswithantioxidantcapacitiesbeinghigherthanthenormal.Inthiscontext,then,thebarberryisasuit-able plant for relevant investigation. Formulatingmilk by barberryjuice, pectin, sugar, and producing acidified milk drinks not onlyincreasesthenutritionalandpharmaceuticalpropertiesofmilkbutalsohasthepotential to increasesalesandpromotetheappealofmilk. Furthermore, it can be applied in different products such asin confectionaries and ice creamproductsor beused as anaturalcolor as an alternative to artificial food coloring.A large range ofdrinks from thoseareprepared from fermentedmilkwith stabiliz-ersaddedtothosepreparedbydirectacidificationwithfruitjuices

2  |     TAHSIRI eT Al.

and/oracidswhichrenderacidifiedmilkdrinks.Heetal.investigatedtheeffectofpHadjustmentandthermaltreatmentontheantioxi-dantcapacityoffruitjuicebeveragesandresultsrevealedthatpas-teurization (63°C/30min) and pH adjustment (pH 3.7 or 6.8) hadeithernon-significantorslighteffectsonfruitjuicemilkbeverage’santioxidant capacity (He,Yuan, Zeng,Tao,&Chen, 2015).ThepHofacidifiedmilkdrinks(AMDs)rangefrom3.6to4.2andthiscouldbe accompaniedby sedimentationproblemsand subsequentmac-roscopicwheyseparationinthispHrange.Relevanttothiscontext,Janhøj etal. produced drinking yoghurt made from fruit concen-trateandreconstitutedmilkpowder,andthenthedrinksweresta-bilizedwithpectinand/orcarboxymethylcellulose(CMC)(0–0.5%)(Janhøj,Frøst,& Ipsen,2008).Caseinsareassumed tobemicellesattheneutralpHinmilk.Caseinsremaininsuspensionformduetothe hairy layer of k-caseinwhich provides steric and electrostaticrepulsive interactions between caseinmicelles.These interactionscausecaseinstostayintheirsuspendedstate(DeKruif,1998;Holt,1992;Schmittetal.,2000).However,thismechanismofstabilizationofcaseinmicellesfailstobemaintained inthepHvaluearound4,duetothecollapseoftheextendedconformationofk-casein.Thetendencytoincreaseentropyofk-caseinchainscausestherepulsiveinteractionbetweenk-caseinchainsbecausethek-caseinchainsofneighboringmicelles tend to overlap andwould result in the lossof entropyof the chains.This phenomenon is called steric stabili-zation(Tuinier,Rolin,&DeKruif,2002).Indiluteacidifiedmilksys-tems, pectinwas added toAMDswith less than1% (w/w) nonfatmilk solids as the stabilizer. Itwas shown that pectin is adsorbedontothecaseinmicellesbecauseofelectrostaticinteraction(Glahn&Rolin,1994).Appearanceistheforemostcriterionthatinfluencestheacceptanceorrejectionoffoodbyconsumers;therefore,stabi-lizersarewidelyusedtostabilizethesedrinkstopreventthefloccu-lationofmilkprotein.Thisassistsintheattainmentofoptimalmouthfeel,therebyenhancingthefavorablefeaturesofproducts(DeKruif&Tuinier, 2001; Glahn & Rolin, 1994). Sensory qualities of foodscanbeevaluatedbasedonestimationsof thetotal impressionthefoodmakesonthemindofthepersonconsumingthefood(Giusti,Bignetti,&Cannella,2008;Reinoso,Mittal,&Lim,2008).Thesen-soryevaluationoffoodisoftenregardedasbeingcharacterizedbyinaccuracy,mistakennessanduncertain repeatability.Nonetheless,sensory data,viz. appearance, taste,mouth feel, aroma, and colorarenormallyanalyzedstatistically,andyetit isnotpossibletofindout precisely by such analyses the strength andweaknessof spe-cificsensoryattributeswhicharemainlyinvolvedindeterminingtheacceptanceorrejectionofthedrinks.Thisshowstheimportanceofsuchdecision-makingasa tool for ranking thequalityofproductsevaluatedbythepanelists.Itisalsoamethodwhichaidsincompar-ingnewproductswithsimilarproductsalreadyinmarkets.Accordingtoearlierreportsbyresearchers,theFuzzylogicisausefultoolthatcanbeemployedwhenconductinganalysesonsensorydataofmanyfoodproductslikedrinks(Lazim&Suriani,2009).

Thepresentarticlewasundertakentoelucidatetheacceptablelevelofingredientswhichcanbeincorporatedintotheproductionofmilkbarberryjuicedrinkwiththegreateststability.Theobjective

isachievedbyanalyzingevaluationsonphysicochemicaldata.Thisarticlealsoaimsto investigatethequalityofproducedAMDsam-plesthroughsensoryevaluationbyrankingtheAMDsampleswithrespect to their quality attributes, using the fuzzy logic.Attemptswere also made to find out the strength and weakness of eachsample.

2  | MATERIALS AND METHODS

AllchemicalswereofanalyticalgradeandwerepurchasedeitherfromMerck (Darmstad,Germany)orSigma–Aldrich (St.Louis,MO,USA).Inordertopreparebarberryjuice(20°Bx),seedlessbarberry(Berberis vulgaris)waspurchased froma localmarket in Shiraz, Iran. Low fatmilkwaspurchasedfromPegahFarsDairyCompany(Shiraz,Iran).ToformulatetheAMDs,GENU®PectintypeYM-150-L(withanesteri-ficationdegreeof72%)waspurchasedfromCPKelco(LilleSkensved,Denmark).

2.1 | Preparation of barberry juice concentrate

Thestalksofseedlessbarberry(Berberis vulgaris)weredetachedfromthe fruitswhichwere then stored in cold storageat−18°C.Beforeextractingthejuice,thebarberrieswerewashed.Thejuicewasmixedwith distilledwater (5: 1). Thismixturewas then exposed to ultra-sound Bandelin (DT255 H, BANDELIN electronic GmbH and Co,Berlin,Germany)at35°C.Thebarberryjuicewassubsequentlytrans-ferred toaKenwoodblender (Blend-XClassicBLP607WH,Havant,England)andwasmixedfor5minat1,000rpm.Finally,themixturewasconcentratedinarotaryevaporatorat30°C(ModelR-3,BUCHICo,Flawil,Switzerland)toreach20°Bx.

2.2 | Chemical analysis of barberry concentrate

2.2.1 | Antioxidant activity

Inthisstudy,thetotalantioxidantactivity,anthocyaninscontent,andtotalphenoliccompoundofthebarberryjuiceconcentrate(BJC)wasdetermined.

Freeradicalscavengingactivity(RSA)oftheBJCwasdeterminedaccordingtothemethodofCam,Hisil,andDurmaz(2009).Avolumeof0.1mlofsampleswasmixedwith0.9mlof100mmol/LTris-HClbuffertowhich1mMofDPPHwasaddedandvortexed.Thereactionmixturewasleftinthedarkfor30min,afterwhichtheabsorbanceoftheresultingsolutionofBJCwasrecordedbyaReyleighspectropho-tometer (ModelVIS-7220G/UV9200, Beijing Beifen-Ruili Analyticalinstrument(Group)Co,Beijing,China)at517nm.Thecontrolsamplewaspreparedinasimilarwaybyadding0.1mlofwaterinsteadofjuicesample.Theantioxidantactivitywasexpressed inthepercentageofinhibitionoftheDPPHradical,andwasdeterminedbythefollowingequation:

(1)RSA%=[(Acontrol−Asample)∕Acontrol]×100

     |  3TAHSIRI eT Al.

where Asampleisthebarberryconcentrateabsorbanceat517nmandAcontrol is the control sample absorbance at 517nm (Çam, Hışıl, &Durmaz,2009).

2.2.2 | Total anthocyanin measurement

ThetotalanthocyanincontentofBJCwasdeterminedbythepHdif-ferential method based on structural transformations of anthocya-ninsasafunctionofpH-generatingcolorsolutions.Theflaviliccationexhibits redcolorand is themostprominent formof its type inpH1.0,whilecarbinoliscolorlessandismostabundantatpH4.5.Inthismethod,twobuffersystemsareused:thepotassiumchloridebufferpH1.0(0.025mol/L)andthesodiumacetatebufferpH4.5(0.4mol/L)(Cametal.,2009).Briefly,0.4mloftheBJCsamplewasmixedwith3.6mlofthecorrespondingbufferandwasreadagainstwater,astheblank,at510nm(A510)and700nm(A700).Absorbancewasdeter-minedbythefollowingequation:

Total anthocyanins content (TAC) of the sample (mg cyanidin-3-glucoside/100mlofBJC)wascalculatedbythefollowingformula:

whereAistheabsorbance,MWisthemolecularweight(449.2),DFisthedilutionfactor(10),andMAisthemolarabsorptivityofcyanidin-3-glucoside(26.900).

2.2.3 | Total phenolic content measurement

Totalphenoliccontent inextractswasdeterminedbyFolin–Ciocalteucolorimetric method according to Vinson etal. (1995); 0.5ml of theconcentratedjuicesamplewasmixedwith2.5mlfolin–ciocaltieuasthechemicalreagent(0.2N)andtheresultantsolutionwaspouredintoeachtube.Thesolutions insidethetubesweremixedbyashakerfor30s,andthen2mlofsodiumcarbonate7.5%wasaddedafter3minofrest.Thesolutionwasfurthermixedfor30satambienttemperaturefor1hr.Absorptionofeachsamplewasreadat765nmbyaspectrophotometer.Theamountofphenolcomponentscanbecalculatedbythecalibrationofgallicacidasthestandard(Vinson,Dabbagh,Serry,&Jang,1995).

2.2.4 | Vitamin C measurement

ThevitaminCwasquantifiedusingtheKnauer1000high-performanceliquidchromatography(HPLC)(KnauerCorp.,Berlin,Germany),whichwas equipped with Nucleodur, C18 pyramid (250×4.6mm, 5μm,Germany)fittedwiththesameguardcolumn.Agradientofmobilephasecreatedofmethanol(solventA)and5mmol/lKH2PO4,pH2.65(solventB)wasusedaccordingtothefollowingprogram:linearincrementstart-ingwith 5%–22%A in 6min and the return to the initial conditionswithinthenext9minwiththeflowrateof0.8ml/min.Theeluatewasdetected,usingaKnauer2600photodiodearraydetectorsetat245nm(Gliszczyńska-Świgłoetal.,2006).Theinjectionvolumewas20μL.

2.3 | Preparation of milk- barberry drink

AMDswerepreparedbymixingthemilk-barberry(Milk:Barberry)inratiosof9:1,8:2,7:3,6:4,5:5,6:4withhighmethoxylpectinsolu-tionsatconcentrationsof0.2%,0.3%,0.4%and0.5%.Accordingly,twenty-fourformulaeweretested ina6×4design intwo levelsofhomogenized and nonhomogenized solutions (with two variables:Milk:Barberryratioandpectinconcentration).Forthispurpose,0.2%,0.3%,0.4%and0.5%pectinsolutionswerepreparedfromhighmeth-oxyl pectin (HMP) at 75°C andwere stored overnight at 4°C. Theproperamountofmilkwastemperedina40°Cwaterbathaddedtothepectinsolutionsandwasmixedbyrotatingat500rpmfor5minon themagneticstirrer.Then, thespecifiedamountofBJCand8%sugarwereadded.Allsampleswerestirredandsubsequentlyhomog-enizedat10,200rpmfor2.5min(IKAT25digitalULTRA-TURRAX)andwerethenpasteurizedat72°Cfor15sandstoredinsealedplas-tic centrifuge tubes at 4°C for 10days. Table1 depicts all the for-mulations related to the production ofmilk-barberry concentrationbeverage.(2)A= (A510−A700)pH1.0− (A510−A700)pH4.5

(3)TAC= (A×MW×DF×100)∕MA

TABLE  1 Specificationofdifferentformulation(a) used to determinethebestformulaforproducedmilk-barberrydrinks

Formulation code

Ratio milk to barberry juice concentrate

Barberry juice concentrate, g Milk, g Pectin, g

S1 9:1 9.18 82.62 0.20

S2 2:8 18.36 73.44 0.20

S3 7:3 27.54 64.26 0.20

S4 6:4 36.72 55.08 0.20

S5 5:5 45.90 45.90 0.20

S6 4:6 55.81 36.72 0.20

S7 9:1 9.17 82.53 0.30

S8 2:8 18.34 73.36 0.30

S9 7:3 27.51 64.19 0.30

S10 6:4 36.68 55.02 0.30

S11 5:5 45.85 45.85 0.30

S12 4:6 55.02 36.68 0.30

S13 9:1 9.16 82.44 0.40

S14 2:8 18.32 73.28 0.40

S15 7:3 27.48 64.12 0.40

S16 6:4 36.64 54.96 0.40

S17 5:5 45.80 45.80 0.40

S18 4:6 54.96 36.64 0.40

S19 9:1 9.16 82.44 0.50

S20 2:8 18.32 73.28 0.50

S21 7:3 27.48 64.12 0.50

S22 6:4 36.64 54.96 0.50

S23 5:5 45.80 45.80 0.50

S24 4:6 54.96 36.64 0.50

aItisnecessarytomentionthatinallformulations,fixedamountofsugarwasconsidered(i.e.,equalto8g).

4  |     TAHSIRI eT Al.

2.4 | Physicochemical properties of milk- barberry drink

2.4.1 | PH measurements

The pH of samples were monitored immediately after production (pH/mV,modelPT370,Keison,UK).

2.4.2 | Total solid measurements

Total amount of dry solid is determined by the equation TS=100- w.b.(%)(Ahmed,Ramaswamy,&Khan,2005).

2.4.3 | Stability measurement

Tomonitor stability, 50ml plastic centrifuge tubeswere filledwithsamplesandwerestoredat4°C.After10daysofstorage,theweightof the layerofwhey, if present,wasmeasuredand thepercentage(w/w)ofwheyseparationwascalculated.Ifthepercentageofwheyseparationismorethan10%,thesampleisconsideredunstable.

2.4.4 | Viscosity measurement

Viscosityofsampleswasevaluated,usingtheBrookfieldViscometer(RVDV-IIPro,BrookfieldEngineering,Massachusetts,USA),equippedwithspindleHA3runningat10,30,60and100rpm.Measurementsweretakenat4°C.

2.4.5 | Color measurement

The color of samples was measured, using digital imaging and thePhotoshop software (Adobe system Inc., san Joes, California, USA)(Afshari-Jouybari&Farahnaky,2011).

2.5 | Statistical analysis

Alldata(exceptsensorydatathatanalyzedbyfuzzylogic)weresta-tisticallyanalyzed,usingtheanalysisofvariance(ANOVA)procedureoftheMinitab16(,StateCollege,PA,USA).Theanalysiswascarriedoutinthreereplications.Tukey’smultiplerangetestswereappliedtodeterminesignificanceofdifferencesbetweenmeanvalues(p<.05).

2.6 | Sensory analysis

Initially,sixmilkbarberrydrinksampleswereevaluatedbasedonthedesirabilityvaluesofpH,andsensoryevaluationwasappliedviaSerumSeparation.Apaneloftwentyhealthypanelists(elevenfemalesandninemales)wasselectedfromamongthestaffmembersandstudentsoftheFoodScienceandTechnologyDepartment,ShirazUniversity,inordertoassesssuccessivemilk-barberrydrinks.Initially,thepanelwasfamiliarizedwiththevariousterminologyemployedtodescribethesensoryevaluation(color,taste,aroma,andmouthfeel),andthescoresheetandmethodofscoring.Thepanelistswereaskedtogive

checkmarks (⋅) to the appropriate respective fuzzy scale factor foreachsampleafterevaluatingthemilk-barberrycomprehensively.Thejudgmentsweretobemadequicklybutnothurriedly.Thepanelistswereaskedtorinsetheirmouthwithwateraftertastingeachmilk-barberry sample. The sampleswere rated as “Poor,” “Fair,” “Good,”“VeryGood”or “Excellent.Thesetsofobservationswereanalyzed,using Fuzzy analysis of sensory scores (Jaya & Das, 2003;Meena,Gupta,Khetra,&Raghu,2015;Routray&Mishra,2012).

2.6.1 | Fuzzy analysis of sensory data

This method utilizes linguistic data obtained by sensory evaluation.Rankingof themilk-barberry sampleswasperformedusing the trian-gular fuzzy membership distribution function as described by SinijaandMishra (2011).Sensoryscoresofthemilk-barberrysampleswereobtained,usingthefuzzyscoresprovidedbythepanelists,whichwereconverted to triplets and used for the estimation of similarity valuesemployed in the rankingof samples. Themajor steps involved in thefuzzymodelingofsensoryevaluationwere:(1)calculationofoverallsen-soryscorestripletsofthemilk-barberry;(2)computationofmembershipfunctiononstandardfuzzyscale;(3)estimationofoverallmembershipfunctiononstandardfuzzyscale;(4)estimationofsimilarityvaluesandrankingofmilk-barberrysamples.AprograminMatlab2015a(TheMathWorks)wasdevelopedforthecalculationofall theabove-mentionedsteps. Triangularmembership functiondistributionpatternof 5-pointsensory scalesweredelineatedby a setof threenumbers, knownasthe“triplet”.Thedistributionpatternof5-pointsensoryscalesarecom-posedof“Poor,(0,0,25)”,“Fair,(25,25,25),”“Good,(50,25,25)”“VeryGood(75,25,25)”and“Excellent(100,25,0)”.Thefirstnumberofthethreenumbersshownintheparenthesesdenotesthecoordinateoftheabscissawherethevalueofthemembershipfunctionis1(Figure1),andthesecondandthirdnumbersofthetripletdesignatethedistancetotheleftandright,respectively,ofthefirstnumber,wherethemembershipfunctioniszero(Sinija&Mishra,2011).

2.6.2 | Triplets for sensory scores of milk- barberry drinks and overall quality

Thetriplet(threenumberset)forsensoryscoresofeachqualityitemsofeverysamplecanbecalculatedbythefollowingequation:

where(n1+n2+n3+n4+n5)resemblesthetotalnumberofpan-elists,C is forthecolor item,thesubscript“r” is forsamplenumberwhile n1, n2, n3, n4 and n5 are the number of panelistswho give“Poor”,“Fair”,“Good”,“VeryGood”or“Excellent”scorestoeachqualityattributeofeachsample,respectively.

After calculating the triplets foreachquality attributeof the sixmilk-barberry drinks, the tripletswere incorporated into Equation5thustodeterminetherelativeweightingofsensoryscorespertainingtoeachsensoryitem:

(4)

SrC=n1(0 0 25)+n2(25 2525)+n3(50 2525)+n4(75 02525)+n5(100250)

n1+n2+n3+n4+n5

,

     |  5TAHSIRI eT Al.

whereCisforcolor,Tisfortaste,Aisforaroma,Misformouthfeel,thesubscript“r”isforsamplenumber,QCrel,QArel,QTrel,andQMrel representthetripletsassociatedwiththerelativeweightingofqualitycharacteristicsofmilk-barberrydrinksingeneral.Forexample,inthecaseofthecoloritem,QCrel=SC/Qsum,whereQsumisthesumofthefirstdigitofthetriplets(Routray&Mishra,2012).

2.6.3 | Assessment of membership function for standard fuzzy scale

MembershipvaluesofMembershipfunctionforeachtriangulardistribu-tionpatternofa6-pointscale(illustratedandnamedF1,F2,F3,F4,F5,andF6)weredefinedbyasetof10numbersasshowninEquation(6).

2.6.4 | Estimation of overall membership function of sensory scores on standard fuzzy logic scale

Overallmembershipfunctionvalueofvarioussampleswascalculated,usingoneofthefollowingequations:

whereBxisthevalueofmembershipfunctionofsensoryscoresthatareestimatingatx=0to100;thena,bandcaremembershipval-uesofthesensoryoverallscoresforthetripletofeachsample.Thex

valuecanbeincludedintheformofasetwithtennumbersstartingfrom0<x<10 to90<x<100with intervalsof10,whereby themaximumvaluesofBxoccurredinthementionedrangeofx(Sinija&Mishra,2011).

2.6.5 | Estimation of similarity values and the ranking of the milk- barberry- based drinks

Afterobtaining theBvalues foreachof thesamplesona standardfuzzyscale(Equation7),thesimilarityvaluesofeachindividualsamplewasobtainedbyEquation(8):

whereSmrepresentsthesimilarityvalueofaparticularsample,theF’andB’representthetransposeofmatrixFandB,respectively.

After determining the various similarity values, theywere com-paredwitheachothertofindoutthemaximumsimilarityvalueofeachsampleonsixcategories(viz.notatallnecessary,extremelyimport-antand,etc.)ofsensorialscalesand,accordingly,allsixmilk-barberrydrinkswereranked.Matlab2015a(TheMathworksInc.,Natick,MA)was used for the fuzzy logic analysis of the sensory data (Sinija &Mishra,2011).

3  | RESULTS AND DISCUSSIONS

3.1 | Chemical properties of barberry concentrate

PreliminarytestsconductedontheBJCrevealeditssignificantanti-oxidant activity. Different anthocyanins constitute the majority ofpigments in the barberry fruit. In this study, theBJC’s anthocyanincontentwas evaluatedby thepHDifferentialMethodon thebasisof barberries’ dominant anthocyanin. The total anthocyanins were150.90±0.02 (mg cyanidin-3-glucoside/100ml). The amounts ofanthocyaninswereobservedtoincreaseafterconcentratingthejuicesolution.Thisisduetocopigmentation,copolymerization,andacyla-tion.Totalphenoliccontents,antioxidantactivityandvitaminCwere282.25±0.02 (mgGAE/100ml), 85.90±0.67 (%) and69.55±0.20(mg/L),respectively.

(5)SrO=SrC×QCrel+SrT×QTrel+SrA×QArel+SrM×QMrel

(6)

F1= (1 0.5 0 0000000)

F2= (0.5 1 10.5 0 00000)

F3= (0 0 0.5 1 10.5 0 000)

F4= (0 0 000.5 1 10.5 0 0)

F5= (0 0 00000.5 110.5)

F6= (0 0 0000000.5 1)

(7)

Bx=x− (a−b)

b, for (a−b)<x<a

Bx=(a+c)−x

c, for a<x< (a+c)

Bx=1 for x=a

Bx=0 for all other values of x

,

(8)Sm(F,B)=F×B�

Max(F×F�andB×B�),

F IGURE  1 Triangularmembershipfunctiondistributionpatternof5-pointscale(Jaya&Das,2003;Sinija&Mishra,2011)

6  |     TAHSIRI eT Al.

Amongnaturalantioxidants,phenoliccompoundsareagroupofspecial interestbecauseof theirwidedistribution in theplantking-dom.All thephenolic classes (including the simplephenolics, flavo-noids,phenolicacids,andanthocyanins)havethestructuralrequire-mentsof free radical scavengers andhave thepotential tobeusedasfoodantioxidants(Sun,Chu,Wu,&Liu,2002).Furthermore,vita-minC is agoodcontributor to the total antioxidant capacityof thefood(Chaovanalikit&Wrolstad,2004).ThechemicalanalysisofBJCrevealedtheexistenceofconsiderablelevelsofantioxidantactivityinthisstudy,which is inagreementwithapreviousreportbyHanachietal.regardingrelevantresultsontheantioxidantactivityofB. vulgaris fruits(Hanachi,Kua,Asmah,Motalleb,&Fauziah,2006).

3.2 | Physiochemical properties of milk- barberry drinks

3.2.1 | pH, dry solid and serum separation

From this study, the best formulationwas optimized based on pH(optimizedvaluesof3.5–4.2),drysolid(>%14)andserumseparation

(Table2).Resultsshowedthatamong24formulations,theS19-S24wereabletocreategelstrengthenednetworks,andtheformulationscomingS6,S9,S10,S11,andS12exhibitednonappropriatepropertiesdue toprecipitation (>%10)after10days.Other formulas includingtheS1,S2,S3,S7,S8,S13,S14,andS19werenotintheoptimizedpHrange(3.5–4.2)atthebeginningofproductionandthuswereremovedfromtheexperiment.Therefore,justtheS3,S4,S5,S16,S17,andS18formulationswerechosenamongthe24formulaefortheapplicationofsensoryevaluationsbasedondesirabilityvaluesofpH,serumsepa-rationandviscosity.

Asitwasmentionedpreviously,theinteractionbetweenbiopoly-mershappeneddependingondifferentfactors,amongwhichpHwasthemostimportant.Differentmechanismsincludingthermodynamiccompatibility and thermodynamic incompatibility result in serumseparation.Forthesamplecontainingthemilk-barberrywitharatioof (9:10), themost importantdestabilizationmechanism is thermo-dynamicincompatibilityduetohavinghighpHvalues.FortheothersampleshavinglowervaluesofpH,theelectrostaticinteractionoccursbetweenmilkproteinandpectin;therefore,ahighermixingratio(hav-ingahigheramountofpectin)resultsinmorestability.Avalueof0.2

TABLE  2 Physiochemicalproperties(in3replicates)ofdifferentpreparedformulationinordertoselectthebestformulaofdrink

Formulation code pH Total solid (%)

Serum separation

After homogenization Before homogenization

S1 5.60±0.00a 18.53±0.01s 0.00±0.00k 0.50±0.06k

S2 4.63±0.00c 19.22±0.01o 6.00±0.02h 7.00±00.05h

S3 4.31±0.00d 19.93±0.02l 8.10±0.31f 9.11±0.01fg

S4 3.82±0.00e 20.65±0.02i 8.30±0.35f 8.50±0.10f

S5 3.65±0.00f 21.36±0.02f 7.00±0.10g 10.00±0.00g

S6 3.59±0.00fg 22.14±0.01c 11.00±0.05c 12.00±0.50c

S7 5.59±0.01a 18.62±0.01r 18.00±0.10a 19.00±0.00a

S8 4.61±0.00c 19.29±0.01n 7.50±0.06g 8.50±0.02g

S9 4.31±0.00d 20.00±0.02k 10.00±0.10d 11.5±0.50de

S10 3.81±0.01e 20.71±0.01hi 11.00±0.07c 13.00±0.00c

S11 3.60±0.00fg 21.43±0.01e 10.00±0.03d 11.00±0.07d

S12 3.51±0.01hi 22.21±0.01b 14.00±0.25b 16.00±0.00b

S13 5.57±0.00ab 18.71±0.02q 10.00±0.11d 1.0000.01d

S14 4.60±0.00c 19.35±0.02mn 2.00±0.06j 4.00±0.04j

S15 4.30±0.02d 20.07±0.01j 4.00±0.10i 6.10±0.02i

S16 3.81±0.00e 20.77±0.03h 2.00±0.15j 4.00±0.05j

S17 3.58±0.00g 21.51±0.02d 4.10±0.10i 6.00±0.07i

S18 3.50±0.00hi 22.29±0.01a 9.00±00.50e 9.70±0.07e

S19 5.52±0.01b 18.79±0.01p 0.00±0.00k 0.00±0.00k

S20 4.59±0.00c 19.41±0.02m 0.00±0.00k 0.00±0.00k

S21 4.09±0.10d 20.13±0.01j 0.00±0.00k 0.00±0.00k

S22 3.79±0.01e 20.85±0.05g 0.00±0.00k 0.00±0.00k

S23 3.56±0.00gh 21.57±0.02d 0.00±0.00k 0.00±0.00k

S24 3.49±0.01i 22.35±0.02a 0.00±0.00k 0.00±0.06k

Differentlettersineachcolumnindicateasignificantdifference(p<.05).

     |  7TAHSIRI eT Al.

pectinconcentrationforthesamplescontainingapproximately10%BJCwas not enough to impart phase separationvia excludingvol-umeeffects. Increasing thisvalue to0.3%would result in competi-tive hydration between pectin andmilk protein, thereby increasingthe serum separation as a result of thermodynamic incompatibility.However,inthe0.4pectinconcentration,theviscosityofcontinuousphase increased and thereby reduced the serum separation.At the0.5pectinconcentration,asthegelnetworkwasproduced,theserumseparationwasprevented.Sometrendswereapparentinconnectionwith the results of serum separation; serum separation in samplescontaining0.2%pectinwaslessthansampleswhichcontained0.3%pectin(Figure2).Asmentionedearlier,theprotein–pectininteraction,and consequently the stability ofAMDs depend on the concentra-tionand typeofpectinused, theconcentrationofcaseinand ionicstrength,thepH,andthehomogenizationduringprocessing(Glahn,1982;Glahn&Rolin,1994).Probablylessthana0.3%valueofpec-tinconcentrationwasinsufficientfortheelectrostaticinteractiontoformthesolublecomplexwithallcaseinmicelleswhichwereonlypar-tiallycovered,sothattheproteincanbelinkedbypolymerbridges.However, the 0.5% concentrationofHMPat lowpH adsorbs ontocaseinproteinsurfacesastheresultofelectrostaticreactionandpro-teincontactispreventedbysterichindrance;therefore,thisformsaself-supportingnetworkwhichpromotesthestabilityofthecolloidalsystem(Tromp,deKruif,vanEijk,&Rolin,2004).Furthermore,thesta-bilizationmightbecausedbyacombinationofdepletioninteractionsamong thecomplexofpectin/caseinmicellesandapectinnetwork(Trompetal.,2004).Moreover,resultsrevealedthathomogenizationreduced the serumseparation samples throughout thewhole stud-iedperiod.Thismaybeattributedtoparticlesizereductionandtheincrease in pectin/casein bonding via homogenization. Pertinently,numerousstudieshaveshowntheintenseinfluenceofparticlessizeon the stability of acidifiedmilk (Sedlmeyer, Brack, Rademacher, &Kulozik,2004).

3.2.2 | Viscosity

Increasing the rotational rate from 10rpm to 100rpm led to thedecreaseinapparentviscosityofallsamples(formulationS1toS18)and thus resulted in a verified shear thinning behavior of samples.Furthermore,byincreasingthepectinconcentrationandreducingthemilk-barberryratio,theapparentviscosityincreased.Figure3shows

how apparent viscosity of selected formulations were optimizedbasedonpH(optimized3.5–4.2),drysolid(>%14)andserumsepara-tion(Table2).

3.2.3 | Color analysis

Resultsofdifferentvarianceanalysisof thebest formulationswereoptimized based on pH (optimized 3.5–4.2), dry solid (>%14) andserumseparationwithregardtothecolorindex(L*,a*,b*)asshowninTable3.

The highest and the lowest magnitudes of lightness (L*) wereobserved,respectively,intheM:Bratio(7:3)containing0.4%pec-tinandintheM:Bratio(5:5)containing0.2%pectin.Inthecaseofyellownessindex(b*)andrednessindex(a*),thelowestandhighestvalues pertained, respectively, to theM: B ratio (7: 3) containing0.4%pectinandtheM:Bratio(4:6)containing0.4%pectin.Resultsindicatedapositivecorrelationbetween thebarberryconcentrateand a* and b*, due to the existenceof pigments and componentslike carotenoids and anthocyanins which make the color of bar-berry.Onthecontrary,thecorrelationbetweenL*andthebarberryconcentratewasnegative,indicatingthedecreaseinmilkwithhighlightness.

F IGURE  2 Thestabilityofacidifiedmilkdrinkswithvariouspectinconcentrations

F IGURE  3 Thevariationofapparentviscosityofoptimumformulaasfunctionofrotationalspeed;S4:milk-barberryratio(6:4)containing0.2%pectin,S5:milk-barberryratio(5:5)containing0.2%pectin,S15:milk-barberryratio(7:3)containing0.4%pectin,S16:milk-barberryratio(6:4)containing0.4%pectin,S17:milk-barberryratio(5:5)containing0.4%pectin,S18:milk-barberryratio(4:6)containing0.4%pectin.Itisnecessarytomentionthatinallformulations,fixedamountofsugarwasconsidered(i.e.,equalto8g)

8  |     TAHSIRI eT Al.

3.2.4 | Sensory analysis of drinks using fuzzy logic

Thesensoryresponseobtainedfromthepanelgroupofpanelistsforthescreenedmilk-barberrysamples (samplesS3,S4,S15,S16,S17,andS18)arepresentedinTable4. Itcanbeobservedthatthepan-elistsrankedtheS4forcolor,S16fortaste,S15,andS16formouthfeel as VeryGood/Excellent by greater proportions. Sensory scoretripletsofeachsamplewerecalculatedthroughEquation4whichisgiveninTable4.

The sensory responses provided by the panelists, the sensoryscore triplet, and the tripletof thecorresponding relativeweight-ingofeachqualityattributearedisplayedinTable5.Resultsshowthat none of the sensory quality attributes are given the “not atall important”scores.Accordingtothegraphical representationofmembershipfunctionofatriplet(a,b,c),thevalueofmembershipfunction is close to 1when the value “a” is large enough and/orthevalue “c” is small (Figure1).Therefore, the taste is the stron-gest quality ofmilk–barberry samples,while color is theweakest.Theimportantqualityattributesofmilk-barberrysamplesingeneralwereratedaccordingtothefollowingorderofimportance:taste>mouth feel > aroma> color.This result supports findings by sev-eralpreviousinvestigations,whichindicatedthattasteisthemostimportant quality attribute (Fatma, Sharma, Singh, Jha, & Kumar,2016;Meenaetal.,2015).Also,RoutrayandMishra(2012)definedqualityattributesinthedahidrinksaccordingtothefollowingorderof importance:Taste > Flavor >Homogeneity > Color (Routray&Mishra,2012).Therefore,theseresultscanplayasignificantroleintheoptimizationofingredientsthatareintendedtobeincorporatedintothemilk-barberrysamples.

For determination of overall sensorial scores of all samples,Equation5 was used and the results are given in Table4. Overallmembership functionvaluesofvarioussamples (Table6)andvaluesof membership function of standard fuzzy scale (nominated as F1,F2,etc.)wereusedforthedeterminationofsimilarityvaluesofmilk-barberrysamples (Equation8).Similarityvalues forsixmilk-barberrysamples in different scale factors are shown in Table7. Bold-facedtextsandhighlightsshowthehighestsimilarityvalueforeachdrink.

The maximum similarity value obtained under the category“good” were observed in (S17: 0.755047) and (S16: 0.723721).

For sample S4, the highest similarity valuewas under the cate-gory “Very good” and for S3, S15 andS18 thehighest similarityvaluebecamepartof thecategory “Satisfactory”.After thecom-parison of the highest similarity values of samples, the rankingwasorderedoutasfollows:SampleS4(verygood)>SampleS17(good)>SampleS16 (good)>SampleS3 (satisfactory)>SampleS15 (satisfactory) > Sample S18 (satisfactory).Thus, itwas clearthat milk-barberry samples with 40% and 50% barberry extractwerethemostacceptablesampleofdrink inthesetof thesam-ples.Overall,thedrinkspreparedwiththeadditionof0.4%pectinwere better than thosewith 2% pectin sample. The addition ofpectin could enhance the textural properties andmouth feel ofdrinks, and the additionof barberry extract improves the flavor.Drinkswith60%concentrationofbarberryextractweretheleastacceptableamongtheothersixsamples.Theadditionofflavortomilkhasbeenfoundto increasethepopularityofdairyproductsin foodmarkets (Routray&Mishra, 2012).However, the resultsrevealed that as the concentrationofbarberry extract increasedbeyond50%, a sour aftertaste ensued after drinking,whichwasnotacceptablebythemajorityofpanelists.Theclassificationofallsamplesintosatisfactory,goodandverygoodalsoimpliesthatallsampleshaveanacceptablequalityforconsumers.Therefore,thefuzzy analysis of sensory attributes demonstrated a good abilityinrankingthemilk-barberrysamples(Table7).Thistechniquehasbeensuccessfullyusedformangodrinks(Jaya&Das,2003),sau-sage (Lee&Kwon,2007),Dahi-BasedDrinks (Routray&Mishra,2012), bread (Singh, Mishra, & Mishra, 2012) instant green teapowder(Sinija&Mishra,2011)kheermohan(Meenaetal.,2015),BeetrootCandy(Fatmaetal.,2016)andgluten-freepasta(Sakre,Das,&Srivastav,2015).

4  | CONCLUSION

Effects of pectin were investigated on the stability of the milk-barberry drink in the presence and absence of homogenization.Resultsshowthathighmethoxylpectinataconcentrationaround0.4% (w/w) was the best concentration for the stabilization ofmilk-barberry drinks in the pH range of 3.5–4.2 and rheological

Formulations**

Color measurement

L* a* b*

Milk:Barberryratio(6:4),pectin0.2g 33.00±0.00c,§ 32.00±0.05c 22.50±0.40c

Milk:Barberryratio(5:5),pectin0.2g 29.40±0.04d 39.00±0.08b 28.00±0.00b

Milk:Barberryratio(7:3),pectin0.4g 53.25±0.04a 26.27±0.03d 18.00±0.25d

Milk:Barberryratio(6:4),pectin0.4g 39.60±0.04b 32.09±0.09c 22.10±0.10c

Milk:Barberryratio(5:5),pectin0.4g 38.70±0.00b 39.16±0.16b 28.00±0.14b

Milk:Barberryratio(4:6),pectin0.4g 32.50±0.30c 40.15±0.00a 30.00±0.30a

**Itisnecessarytomentionthatinallformulations,fixedamountofsugarwasconsidered(i.e.,equalto8g).§Differentlettersineachcolumnindicateasignificantdifference(p<.05).

TABLE  3 Colormeasurementresultsofoptimizedformulations

     |  9TAHSIRI eT Al.

properties showed that the stabilized samples behaved as shearthinning liquids, with viscosity increasing parallel to the increaseinpectinandbarberry juicecontent.Sensoryevaluationwasper-formed for six screened milk-barberry samples by a panel of 20panelists,andthedatawereanalyzed,usingthefuzzylogicmethod.Itwasfoundthattheimportantqualityattributesofmilk-barberrysamples in general were rated according to the following order:

taste>mouth feel>aroma>color.Overall acceptancevaluesofeach sample were determined by the maximum similarity value.Sample S4 (milk: barberry ratio (6:4) containing 0.2% pectin) andsampleS3 (Milk-barberryratio(7:3)containing0.2%pectin)havethe highest and the lowest acceptance, respectively. Therefore,Fuzzyanalysisofsensoryattributeshasagoodpotentialforrankingmilk-barberrysamples.

TABLE  4 PanelistsPreferenceforspecificqualitycharacteristicsofmilk-barberrysamplesandtripletsrelatedwithsensoryscores

Sensory attributes of samples

Sensory scale factors and corresponding numerical values

Sensory scores triplet

Poor (0 0 25)a

Fair (25 25 25)

Good (50 25 25)

Very good (75 25 25)

Excellent (100 25 0)

Color/appearance

S3 0 9 9 2 0 (41.252525)

S4 0 0 3 12 5 (77.52518.75)

S15 0 11 9 0 0 (36.252525)

S16 0 0 4 10 6 (77.52517.5)

S17 0 2 11 5 2 (58.752522.5)

S18 0 4 11 4 1 (52.52523.75)

Taste

S3 1 8 6 2 1 (37.521.2521.25)

S4 0 1 9 7 3 (652521.25)

S15 2 7 7 4 0 (41.2522.525)

S16 0 1 8 9 3 (7026.2522.5)

S17 0 6 10 2 2 (502522.5)

S18 3 7 5 2 3 (43.7521.2521.25)

Aroma/Smell

S3 0 5 7 6 2 (56.252522.5)

S4 0 3 9 4 4 (61.252520)

S15 0 5 8 5 2 (552522.5)

S16 0 2 9 6 3 (62.52521.25)

S17 0 3 8 6 3 (61.252521.25)

S18 0 2 9 5 4 (63.752520)

Mouthfeel

S3 0 4 10 6 0 (52.52525)

S4 0 5 8 6 1 (53.752523.75)

S15 0 2 6 8 4 (67.52520)

S16 0 2 6 9 3 (66.52521.25)

S17 0 6 10 3 1 (48.752523.75)

S18 0 7 11 2 0 (43.752525)

Overall

S3 (46.86137.60929.737)

S4 (64.12643.65929.855)

S15 (50.21338.84130.045)

S16 (68.95445.34430.195)

S17 (54.51140.85130.077)

S18 38.78629.561)

aTripletsrelatedwith5-pointsensoryscale.

10  |     TAHSIRI eT Al.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the financial support of theResearchAffairsOfficeatShirazUniversity(GrantNumber–92SU-AG-023).TheauthorsthankMr.MohsenHamedpour-Darabiforfinalproofreadingthemanuscript.

CONFLICT OF INTEREST

Nonedeclared.

REFERENCES

Afshari-Jouybari,H.,&Farahnaky,A.(2011).EvaluationofPhotoshopsoft-warepotentialforfoodcolorimetry.Journal of Food Engineering,106(2),170–175.

Ahmed,J.,Ramaswamy,H.,&Khan,A.(2005).Effectofwateractivityonglasstransitionsofdatepastes.Journal of Food Engineering,66(2),253–258.

Cam,M.,Hisil,Y.,&Durmaz,G.(2009).CharacterisationofpomegranatejuicesfromtencultivarsgrowninTurkey.International Journal of Food Properties,12(2),388–395.

Çam,M.,Hışıl,Y.,&Durmaz,G.(2009).Classificationofeightpomegranatejuicesbasedonantioxidantcapacitymeasuredbyfourmethods.Food Chemistry,112(3),721–726.

Chaovanalikit,A.,&Wrolstad,R.,(2004).Totalanthocyaninsandtotalphe-nolicsoffreshandprocessedcherriesandtheirantioxidantproperties.Journal of Food Science,69(1),FCT67–FCT72.

DeKruif, C. (1998).Supra-aggregates of caseinmicelles as a prelude tocoagulation.Journal of Dairy Science,81(11),3019–3028.

DeKruif,C.,&Tuinier,R.(2001).Polysaccharideproteininteractions.Food Hydrocolloids,15(4),555–563.

Fatma,S.,Sharma,N.,Singh,S.P.,Jha,A.,&Kumar,A.(2016).FuzzyAnalysisofSensoryDataforRankingofBeetrootCandy.International Journal of Food Engineering,2(1).

Giusti, A. M., Bignetti, E., & Cannella, C. (2008). Exploring new fron-tiers in total food quality definition and assessment: From chemicalto neurochemical properties. Food and Bioprocess Technology, 1(2), 130–142.

Glahn,P.,(1982).HydrocolloidstabilizationofproteinsuspensionsatlowpH.Progress in Food and Nutrition Science, 6,171–177.

Glahn,P.,&Rolin,C.,(1994).Casein–pectininteractioninsourmilkbever-ages,FoodingredientsEuropeconferenceproceedings,p.258.

Gliszczyńska-Świgło, A., (2006). Antioxidant activity of water solublevitaminsintheTEAC(troloxequivalentantioxidantcapacity)andthe

TABLE  5 Sumofindividualpreferencetotheimportanceofqualityattributesofsamplesingeneral

NI SI I HI EI Sensory score tripletTriplets for relative weightage

Color/appearance 0 1 3 6 10 (81.252512.5) (0.23550.07250.0362)

Taste 0 0 0 8 12 (902510) (0.26090.07250.0290)

Aroma 0 0 2 8 10 (852512.5) (0.24640.07250.0362)

Mouthfeel 0 0 0 9 11 (88.752511.25) (0.25720.07250.0326)

NI*-notatallimportant,SI*-somewhatimportant,I*-important,HI*-highlyimportant,EI*-extremelyimportant.

TABLE  7 Similarityvaluesformilk-barberrysamples

Sensory scale 1 (S3) 2 (S4) 3 (S15) 4 (S16) 5 (S17) 6 (S18)

F1(NS) 0.050684 0.000000 0.032811 0.000000 0.023200 0.030857

F2(fair) 0.390858 0.129771 0.316583 0.090500 0.264538 0.309766

F3(satisfactory) 0.790112 0.494715 0.730712 0.397695 0.264538 0.730078

F4(good) 0.641480 0.539342 0.692285 0.723721 0.755047 0.704614

F5(verygood) 0.155784 0.725778 0.203074 0.632378 0.456964 0.210306

F6(excellent) 0.000000 0.107751 0.001182 0.174255 0.022898 0.001498

Rank VI I IV III II V

Bold-facedtextsshowthehighestsimilarityvalueforeachdrink;NS:notsatisfactory.

TABLE  6 Overallmembershipfunctionvalueofvarioussamples

Overall membership function Value

B1(3) 0.01990 0.28580 0.55200 0.81760 1.00000 0.89440 0.55818 0.22190 0.0000 0.00000

B2(4) 0.00000 0.00000 0.21840 0.44741 0.67640 0.90550 1.00000 0.80324 0.46830 0.13340

B3(15) 0.00000 0.22213 0.47960 0.73710 0.99450 1.00000 0.67420 0.34140 0.00860 0.00000

B4(16) 0.00000 0.00000 0.14090 0.36145 0.58200 0.80250 1.00000 0.96530 0.63420 0.30300

B5(17) 0.00000 0.15520 0.40000 0.64480 0.88960 1.00000 0.81750 0.48500 0.15250 0.00000

B6(18) 0.00000 0.20750 0.46530 0.72310 0.98100 1.00000 0.68670 0.34840 0.01010 0.00000

     |  11TAHSIRI eT Al.

FRAP (ferric reducingantioxidantpower) assays.Food Chemistry, 96,131–136.

Hanachi,P.,Kua,S.,Asmah,R.,Motalleb,G.,&Fauziah,O.(2006).CytotoxiceffectofBerberis vulgarisfruitextractontheproliferationofhumanlivercancer cell line (HepG2) and its antioxidant properties. International Journal of Cancer Research,2,1–9.

Harborne, J. B. (2013). The flavonoids: Advances in research since 1980. Michigan,USA:Springer.

He,Z.,Yuan,B.,Zeng,M.,Tao,G.,&Chen,J. (2015).Effectofsimulatedprocessingon theantioxidantcapacityand invitroproteindigestionof fruit juice-milk beverage model systems. Food Chemistry, 175,457–464.

Holt,C.(1992).Structureandstabilityofbovinecaseinmicelles.Advances in Protein Chemistry,43,63–151.

Janhøj, T., Frøst, M. B., & Ipsen, R. (2008). Sensory and rheologicalcharacterization of acidified milk drinks. Food Hydrocolloids, 22(5),798–806.

Jaya,S.,&Das,H.(2003).Sensoryevaluationofmangodrinksusingfuzzylogic.Journal of Sensory Studies,18(2),163–176.

Lazim,M.,&Suriani,M.(2009).Sensoryevaluationoftheselectedcoffeeproductsusingfuzzyapproach.World Academy of Science, Engineering and TechnologyTechnol,50,717–720.

Lee,S.J.,&Kwon,Y.A. (2007).Studyonfuzzyreasoningapplicationforsensoryevaluationofsausages.Food Control,18(7),811–816.

Meena, G. S., Gupta, V. K., Khetra, Y., & Raghu, H. (2015). Fuzzylogic (Similarity Analysis) modelling for sensory evaluation of themarket samples of Kheer Mohan. Indian Journal of Dairy Science,68(4).

Reinoso,E.,Mittal,G.S.,&Lim,L.-T.(2008).Influenceofwheyproteincom-positecoatingsonplum(PrunusdomesticaL.) fruitquality.Food and Bioprocess Technology,1(4),314–325.

Rodríguez-Roque,M. J., Rojas-Graü,M.A., Elez-Martínez, P., &Martín-Belloso, O. (2013). Changes in vitamin C, phenolic, and carot-enoid profiles throughout in vitro gastrointestinal digestion of ablended fruit juice. Journal of Agricultural and Food Chemistry, 61(8), 1859–1867.

Routray,W.,&Mishra,H. (2012). Sensory evaluationof different drinksformulatedfromdahi(indianyogurt)powderusingfuzzylogic.Journal of Food Processing and Preservation,36(1),1–10.

Sakre,N.,Das,A.B.,&Srivastav,P.P. (2015). FuzzyLogicApproach forProcessOptimizationofGluten-FreePasta.Journal of Food Processing and Preservation,40(5),840–849.

Schmitt,C.,Sanchez,C.,Despond,S.,Renard,D.,Thomas,F.,&Hardy,J.(2000).Effectofproteinaggregatesonthecomplexcoacervationbe-tween β-lactoglobulin and acacia gumat pH4.2.Food Hydrocolloids,14(4),403–413.

Sedlmeyer,F.,Brack,M.,Rademacher,B.,&Kulozik,U. (2004).Effectofproteincompositionandhomogenisationonthestabilityofacidifiedmilkdrinks.International Dairy Journal,14(4),331–336.

Singh,K.,Mishra,A.,&Mishra,H.(2012).Fuzzyanalysisofsensoryattri-butesofbreadpreparedfrommillet-basedcompositeflours.LWT- Food Science and Technology,48(2),276–282.

Sinija,V.,&Mishra,H. (2011).Fuzzyanalysisof sensorydata forqualityevaluationandrankingofinstantgreenteapowderandgranules.Food and Bioprocess Technology,4(3),408–416.

Slattery,M.L.,Benson,J.,Curtin,K.,Ma,K.-N.,Schaeffer,D.,&Potter,J.D.(2000).Carotenoidsandcoloncancer.The American Journal of Clinical Nutrition,71(2),575–582.

Sun,J.,Chu,Y.-F.,Wu,X.,&Liu,R.H. (2002).Antioxidantandantiprolif-erative activities of common fruits. Journal of Agricultural and Food Chemistry,50(25),7449–7454.

Tromp, R. H., de Kruif, C. G., van Eijk, M., & Rolin, C. (2004). On themechanism of stabilisation of acidified milk drinks by pectin. Food Hydrocolloids,18(4),565–572.

Tuinier,R.,Rolin,C.,&DeKruif,C.(2002).Electrosorptionofpectinontocaseinmicelles.Biomacromolecules,3(3),632–638.

Vinson,J.A.,Dabbagh,Y.A.,Serry,M.M.,&Jang,J. (1995).Plantflavo-noids, especially teaflavonols, arepowerful antioxidantsusing an invitrooxidationmodelforheartdisease.Journal of Agricultural and Food Chemistry,43(11),2800–2802.

How to cite this article:TahsiriZ,NiakousariM,Khoshnoudi-NiaS,andHosseiniSMH.Sensoryevaluationofselectedformulatedmilkbarberrydrinksusingfuzzyapproach.Food Sci Nutr.2017;00:1–11.doi:10.1002/fsn3.454