Resource integration optimization of convenience service ...

Resource integration is the process of effectively identifying, acquiring, and allocating external resources (Ma et al. 2019), and is a critical ... Skiptomaincontent Advertisement SearchallSpringerOpenarticles Search DownloadPDF Research OpenAccess Published:18March2021 Resourceintegrationoptimizationofconvenienceserviceplatformsadoptingdynamicservicemodesinnewretail YangYang1&JianmingYao1  FrontiersofBusinessResearchinChina volume 15,Article number: 2(2021) Citethisarticle 1962Accesses 1Citations Metricsdetails AbstractThispaperanalyzesthekeyissuesofresourceintegration,andproposesamethodforsolvingtheproblemofachievingefficientresourceintegrationforconvenienceserviceplatformsthatadoptdynamicservicemodes.Thismethodconsidersboththeservicelevelofcollaborativeresourcesunderdynamicservicemodes,andcollaborativeresources’satisfactionofservicemodesintheresourceintegration.Inaddition,dominantfactorsneedingthemostattentionduringresourceintegrationunderdynamicservicemodesareexcavated.Moreover,amodelforresourceintegrationoptimizationisestablished,andantcolonyalgorithmisusedtosolvetheproblem.Finally,acaseisusedtoverifythefeasibilityofthemethod. IntroductionWiththeacceleratingpaceoflife,customersareseekingfurtherconvenienceofretailservice.Atpresent,conveniencestores,avitalformofretailservice,havebeenincreasinglyfavoredbyconsumersfortheirconvenience,rapidity,andaccessibilityinanemergency.Thepresenceofconveniencestoreshasbecomeanessentialpartofdailylife(Sakashita2000).Thelargestretailgroupshavedevelopednewconveniencestores,thus,theretailmarketshareofconveniencestoresisrising(GahinetandCliquet2018).Moreover,asnovelmodes,likenewretail,whichmergesbrickandmortarstoreswithecommerce,areemerging(WangandLi2019),itisfeasibleforalargenumberofchainconveniencestorestocombinethemulti-endscaleeffectandbrandreputationwithwell-developedonlineoperationstotransformintoconvenienceserviceplatforms.Forinstance,anumberofFreshipposupermarkets,flagshipsofAlibaba,havesetuplifeservicecentersatwhichconvenienceserviceslikelaundry,hairsalonsandrepairscanbefound.ConvenienceBee,acashier-freeconveniencestore,hasalsoofferedlaundryservices,coveringtheneighborhoodsandbusinessdistrictsinthevicinityofitshundredsofstores.SuningXiaodian,theconveniencestoreofChineseretailerSuning.comCo.,nowprovidesvariouslifeservices,includinghousekeepingandhomeappliancecleaning.ThegeneraloperationalframeworkofconveniencestoresaftertransformationisdemonstratedinFig. 1. Fig.1ServiceformofconveniencestoresaftertransformationFullsizeimageAsshowninFig.1,thecenterisaconvenienceserviceplatformtransformedfromaconveniencestorethatoffersavarietyofdailyconvenienceserviceitems,suchascarwashinganddeliveryservices,todifferentcustomersk(k = 1,2,...K).Asaserviceintegrator,convenienceserviceplatformsneedtointegrateappropriateserviceresourcesr(r = 1,2,...R)toprovidetheseservices.Foreachserviceitemm(m = 1,2,...M),theconvenienceserviceplatformcanchoosealternativeservicemodesn(n = 1,2,...N)toprovidetheservices.Ontheonehand,differentservicemodesofferdifferentlevelsofconvenienceandcost;ontheotherhand,differentserviceresourcesofferdifferentservicelevelsunderdifferentservicemodesandhaverespectivepreferencesfordifferentservicemodesgiventheirownability.Apparently,thesuccessofthisplatformdependsonwhethertheintegratedserviceresourcescanprovideconsumerswithasatisfactoryexperience.Tofulfillthegoal,thecoreenterpriseinchargeoftheplatformisexpectedtointegratetheserviceresourcesinaneffectiveandreasonablemanner.Resourceintegrationistheprocessofeffectivelyidentifying,acquiring,andallocatingexternalresources(Maetal.2019),andisacriticalprocessforcompaniestomaximizeresourcefunction(Lyuetal.2019).Inotherwords,theefficiencyofresourceintegrationdeterminestheservicecapabilitiesoftheconvenienceserviceplatforms.Asserviceshavereceivedincreasinglymoreattention,alargenumberofscholarshaveconductedresearchonservicesupplychainmanagementinsectorssuchasoutsourcing(e.g.,Benjaafaretal.2007;Lahirietal.2013),customerservice(e.g.,AlHamadietal.2015;HallandPorteus2000),capacitymanagement(e.g.,CaoandJiang2013;Chingetal.2011),andservicecompetition(e.g.,Shietal.2019;Wangetal.2019).Industriessuchastelecommunications(e.g.,CanakogluandBilgic2007;Hasijaetal.2008),electricity(e.g.,Oliveiraetal.2013)andlogistics(e.g.,Liuetal.2013;LiuandXie2013;Liuetal.2018)havealsobeenexamined.However,theexistingresearchislargelybasedonstaticservicemodes(includingservicetime,delivery,andcontentsofservice);theydonotconsiderthedynamicchangespresentinservicemodes.Inreality,consumers’needswithregardtotheconvenienceofservicesfrequentlyvarywithspecificsurroundingsandscenarios.Therefore,indesigningserviceitemsforconvenienceserviceplatforms,itisadvisabletoconsiderhowtomeetconsumers’changingneedsthroughdynamicservicemodesandexaminehowresourceintegrationcanbeincorporated.Inotherwords,comparedwiththetraditionalsingleservicemode,theinfluenceofdynamicofservicemodesneedtobeconsideredintheresourceintegrationforconvenienceserviceplatforms.Therefore,thisstudyisaimedatexploringtheoptimizationofresourceintegrationunderdynamicservicemodes.Withthepurposeofofferingamethodforresourceintegrationoptimizationofconvenienceserviceplatformsunderdynamicservicemodes,thepaperfirstanalyzesthekeyissuesofresourceintegration,andthenestablishestheoptimizationmodel.Theantcolonyalgorithmisadoptedtofindthesolution,andacaseisanalyzedtoverifythefeasibilityofthemodel.Thenextsection,Section2,willreviewrelevantliteratureinthisaspect.Section3includesthedescriptionofdynamicservicemodesofconvenienceserviceplatforms.Section4focusesontheexcavationofkeyfactorsforresourceintegrationunderdynamicservicemodes.TheoptimizationmodelofresourceintegrationisestablishedinSection5,andtheselectionandadoptionofthealgorithmisdemonstratedinSection6.InSection7,thefeasibilityofthemodelisverifiedthroughacase.Section8presentstheconclusion.LiteraturereviewThereexistsalargeamountofliteratureonthesupplychainmanagementofconveniencestores,includingstudiesonthelocationanddelivery(e.g.,Hübneretal.2016;Sakashita2000;WoodandBrowne2007),andresearchaboutreplenishmentandcategorymanagement(e.g.,Alftanetal.2015;Hanetal.2014).AsChineseconveniencestoresaretransformingandupgradingintoconvenienceserviceplatforms,itisnecessarytotakeintoaccounttheissueofresourceintegrationforservicesuppliersfromtheperspectiveofserviceoperationmanagement.Inaddition,alargeamountofresearchhasbeenconductedontheselectionandallocationofsupplychainresources.Intermsofresourceintegrationinmanufacturing,researchershavepaidmuchattentiontosustainableresourceintegrationandtheresourceintegrationintheservice-orientedtransformationofmanufacturing.Bastletal.(2017)exploretheroleofthemanufacturer–supplierrelationshipinserviceperformancewithinservicetriads.Luthraetal.(2017)proposeaframeworkforsustainablesupplierselectionandevaluationinsupplychains.LambaandSingh(2019)proposeadynamicsupplierselectionmodelconsideringthecarbonemissionsinabigdataenvironment.Ayalaetal.(2019)studythemoderatingroleofservicesuppliersinproductcompanies(servitization).Withregardtotheintegrationofserviceproducts,somescholarsexploredtheresourceselectionandintegrationinthird-partylogisticsservice.Pamucaretal.(2019)provideamethodfortheassessmentofathird-partylogisticsprovider.Lyuetal.(2019)examinetheimpactoflogisticsplatformsandlocationonlogisticsresourceintegrationandoperationalperformance.However,despitethrivingserviceproducts,researchconcerningtheresourceintegrationoftheservicesectorislacking.Inparticular,fewconsiderthedynamicchangesinservicemodes,thoughtheimportanceofresourceintegrationhasbecomeincreasinglypopular.Supplychainmanagementhasbecomeanessentialmethodinaidingenterprisestoimprovetheiroperationflexibilityandacquireexternalresources(Yao2010;YaoandGu2014).Yao(2017)pointsoutthatasupplychainnetworkcomprisesdifferentsupplychainresources,anditsoperationefficiencyisdeterminedbythelevelofsupplychainresourceintegration.Companiescanimprovetheiroperationalperformanceandobtaincompetitiveadvantagethroughresourceintegration(Lyuetal.2019).Therefore,inviewoftheimportanceofresourceintegrationandthelackofresearchabouttheimpactofdynamicservicemodesonresourceintegration,thisresearchattemptstoexploretheoptimizationofserviceresourceintegrationintermsofdynamicservicemodes.Thetheoreticalcontributionsofthispaperaremainlythefollowing.Firstly,giventhatfewstudiesconsidertheinfluenceofdifferentservicemodesintheprocessofresourceintegrationwithregardtotheadoptionofdynamicservicemodes,thepresentresearchquestioneliminatestheexistinggapasmuchaspossible.Secondly,thecharacteristicsofresourceintegrationunderdynamicservicemodesareanalyzedinthispaper.Specifically,twoaspectsareconsidered,namely,serviceleveldifferenceofcollaborativeresourcesunderdifferentservicemodes,andcollaborativeresources’satisfactiontowarddifferentservicemodes.Inaddition,thekeyfactorsofresourceintegrationunderthedynamicservicemodesareexcavatedandanalyzed.Thirdly,theprocedureoftheproposedmethodisstraightforward,andgenerallyapplicabletotheoptimizationofresourceintegrationoftheconvenienceserviceplatformsthatadoptthedynamicservicemodes.DynamicservicemodesofconvenienceserviceplatformsConvenienceserviceplatformsprovideconvenienceservicesagainstcustomerneedsforconvenience.Berryetal.(2002)defineserviceconvenienceasaconsumers’perceptionofminimizedtimeandeffortrequiredtoobtainaservice.Itisworthnotingthatthedegreeofcustomerdemandforconvenienceisnotconstantandoftenchangeswiththesurroundingenvironmentandconditions.Forexample,thesamecustomer’sneedsforconveniencedifferbetweenweekdaysandweekends,andalsobetweenpeakworkhoursandnoon;inaddition,changesintheenvironment,suchasweatherandtemperature,willalsocausechangesincustomerconvenienceneeds.Undertheconditioninwhichcustomershaveinsufficienttimeorenergy,theypaymoreattentiontotheserviceconvenienceandfactorsthatcanreducethetimeandenergyexpendedinsteadoffactorssuchasqualityandprice;conversely,undertheconditioninwhichcustomershavesufficienttimeorenergy,theypaylessattentiontotheserviceconvenienceandfocusmoreonfactorssuchasqualityandprice.However,theimprovementofserviceconvenienceisoftencoupledwithcostincreases.Thus,convenienceserviceplatformsshouldsupplysuitableconveniencebasedonthedynamicserviceconveniencedemandofcustomers.Serviceconvenienceisamultidimensionalconstruct,andshouldbeunderstoodwithinthecontextoftheactivitiesthatconsumersundergointheprocessofpurchasingandusingaservice(Colwelletal.2008).Brown(1990)proposesfivedimensionsfortheconceptofconvenience:time,place,acquisition,use,andexecution.Astheacquisitionandusedimensionsdonotvarywithservicemodesbecausetheyareinfluencedbythetransactionchannelsandpaymentmethodsofconvenienceserviceplatforms,inthisresearch,wemainlyfocusonthetime,place,andexecutiondimensions,whichdovarywithservicemodes.Takingacommunityconveniencestorewithlaundryservicesasanexample,Table 1showsdifferentservicedetailsintermsofthedifferentservicedimensionsthata convenienceserviceplatform canprovide.Theservicecontentschosenbya convenienceserviceplatformagainstthedifferentdimensionsconstitutethewholeservicemodeofthatserviceitem,asshownbytheexampleinTable 2.Obviously,theconveniencelevelsoftheservicemodesaredifferent. Table1ExamplesofdifferentservicedetailsprovidedbyaconvenienceserviceplatformbyconveniencedimensionsFullsizetableTable2ExampleofthedifferentservicemodesofaconvenienceserviceplatformwithlaundryserviceFullsizetableInreality,platformsgenerallycannotprovideallthepossibleservicemodesforcustomersbecausetoohighadegreeofcustomizationinservicemodescannoteffectivelyproduceeconomiesofscale.Therefore,platformsneedtoselectanddesignservicemodesbasedontargetcustomers’conveniencedemand.Asstaticservicemodesfinditdifficulttomeetchangingcustomerdemand,theplatformsoftenneedtosetupdynamicservicemodes.Forexample,asshowninFig.2,aconvenienceserviceplatformcanusethecombinationofservicemodes(3)and(6)forcustomerstochooseinacertainperiod(suchasperiodt),andinanotherperiod(suchast + 1),adoptthecombinationofservicemodes(2)and(4). Fig.2SketchofdynamicservicemodesofconvenienceserviceplatformsFullsizeimageConvenienceserviceplatformscandividetheresourceintegrationdecisionperiodintoTdifferentstagesaccordingtothechangesincustomerconvenienceneeds.Customers’desiredservicemode(obtainedthroughapilotsurvey)canbeusedateachstage.Theselectionratioofeachservicemodeduringtheentireresourceintegrationdecisionperiodcanthenbeobtained.KeyissuesofresourceintegrationofconvenienceserviceplatformsunderdynamicservicemodesInthispart,theproceduresinvolvedinresourceintegrationunderdynamicservicemodesaredefined,andkeyfactorsintheintegrationprocessarediscussed.ResourceintegrationprocedureunderdynamicservicemodesAsmentionedearlier,thekeytothesuccessofconvenienceserviceplatformsisbeingabletomatchcustomerdemandtotheconvenienceexperience,whichreliesontheeffectiveandreasonableintegrationofserviceresources.However,theresourceintegrationprocessofconvenienceserviceplatformsbasedondifferentconvenienceservicemodesisatypicaldynamicprocesswithdifferentcharacteristics.Therefore,theserviceresourceintegrationofconvenienceserviceplatformsshouldbedifferentfromthatadoptedbystaticservicemodes.TheresourceintegrationprocessbasedondynamicservicemodesisdescribedinFig. 3. Fig.3FlowchartofresourceintegrationofconvenienceserviceplatformsunderdynamicservicemodesFullsizeimageAsshowninFig.3,whenconvenienceserviceplatformsunderdynamicservicemodesintegrateresources,theymustfirstclarifytheiradoptionofthedifferentservicemodesusingadynamicdecision-makingprocess.Thedecision-makingperiodfortheresourceintegrationcanbedividedintodifferentstagesaccordingtothechangesincustomers’conveniencedemand.Foreachdelimitedstage,convenienceserviceplatformsselectoneortwoservicemodestomatchcustomerdemand;thus,theconvenienceserviceplatformscandeterminetheadoptionofservicemodesduringthedecision-makingperiodofresourceintegration.Theevaluationofcollaborativeresourcesismorecomplicatedwhenconvenienceserviceplatformsadoptdynamicservicemodesthanwhentheyadoptstaticservicemodes.Specifically,twoaspectsneedtobeconsidered,namely,serviceleveldifferenceofcollaborativeresourceunderthedifferentservicemodes,andcollaborativeresources’satisfactiontowarddifferentservicemodes.Inaddition,notalltheresourcefeaturescanbeconsideredwhenintegratingresources,sothoseneedingthemostattentionshouldbeclarified.Therefore,convenienceserviceplatformsneedtoexcavatethedominantfactorsforresourceintegrationbasedondynamicservicemodes,andcarryoutresourceintegrationinafocusedandtargetedmanner.Convenienceserviceplatformsunderdynamicservicemodesshouldcomprehensivelyconsidertheaboveaspectstooptimizeresourceintegration.Collaborativeresources’servicelevelunderdynamicservicemodesThisarticleassumesthatconvenienceserviceplatformsareonlyresponsibleforfacilitatingtransactions,otherservicestepsarecompletedbytheircarriedserviceresources,andeachcarriedresourcecancooperatewithplatformstoutilizedifferentservicemodesfortheservice.Itisworthnotingthatdifferentserviceresourceshavedifferentservicelevelsunderdifferentservicemodes,giventheirownability.Forexample,theservicesupplier’slogisticalspeedwillaffectitsservicelevelwhenthedoor-to-doordeliverymodeisadopted,andtheserviceprovider’scapabilityforprovidingcustomizedserviceswillaffectitsservicelevelwhenthecustomizedservicemodeisadopted.Thisleadstochangesintheservicelevelofcollaborativeresourceswhenconvenienceserviceplatformsadoptdifferentservicemodes,sotheevaluationofcollaborativeresources’servicelevelshouldbeconsideredincombinationwithdynamicservicemodes.Consideringthatcustomersmayhavedifferentexpectationstowardserviceperformancefordifferentservicemodes,andthatperceivedservicelevelsareinfluencedbythe(dis)confirmationofexpectedserviceperformance(BoltonandDrew1991;Oliver1977;Rustetal.1999;SzymanskiandHenard2001),thecustomerperceivedvalue,insteadoftheobjectivevalue,shouldbeusedtoreflecttheservicelevelofsuppliersunderdifferentservicemodes.Inlinewithprospecttheory(KahnemanandTversky1979)whichbelievespeoplederiveutilityfromgainsandlosses,measuredrelativetothereferencepoint,ratherthanfromtheabsolutelevelsofwealthandaremoresensitivetolossesthantogainsofthesamemagnitude,wetakethecustomerexpectationofservicelevelasthereferencepointtocalculatetheirperceivedvalue.Iftheperformanceoftheactualserviceexceedscustomerexpectations,gainisappreciated,andtheperceivedvalueincreases;iftheperformanceoftheactualserviceislowerthanthatofcustomerexpectations,lossisappreciated,andtheperceivedvaluedecreases.Accordingtothevaluefunctionofprospecttheory,whendescribingthecustomerperceivedvalue,weassumethat,undertheservicemodeAmn,fmnisthefthindexaffectingtheservicelevel,thecustomerexpectationforserviceperformancewiththeindexfmnislfmn,andtheactualperformanceofcollaborationresourceiislf(mn)i.Accordingtothevaluefunctioninprospecttheory,theformulaforthecustomerperceivedvaluevf(mn)ioflf(mn)irelativetolfmnis: $${v}_{f(mn)i}=\left\{\begin{array}{c}{\left({l}_{f(mn)i}-{l}_{fmn}\right)}^{\alpha},{l}_{f(mn)i}\ge{l}_{fmn}\\{}-\uptheta{\left({l}_{fmn}-{l}_{f(mn)i}\right)}^{\beta},{l}_{f(mn)i} 1.Throughtheexperimentaltestsanddataregressionanalysisonalargenumberofdecision-makingindividuals,TverskyandKahneman(1992)concludethatα = β = 0.88andθ = 2.25,whicharetheparametervaluesthatcanapproximatelyindicatethebehavioralpreferencesofthedecision-maker.Therefore,thisarticletakestheabovevaluesastheparameters.Assumethattheimportanceattachedbycustomerstotheindexf(mn)isWf(mn), then,undertheservicemodeAmnadoptedbytheconvenienceserviceplatform,thecustomerperceivedvalueabouttheservicelevelofthecollaborativeresourcei,E(mn)i ,canbeexpressedas: $${E}_{(mn)i}={\sum}_{f=1}^F{W}_{f(mn)}{v}_{f(mn)i}.$$ (2) Combinedwithumn,theproportionofeachservicemodeutilizedbyconvenienceserviceplatformsduringthedecision-makingstageofresourceintegration,thecustomers’overallperceivedvalueofthecollaborativeresourceiunderdynamicservicemodescanbeexpressedas: $${E}_{mi}={\sum}_{n=1}^N\left[{u}_{mn}\left({\sum}_{f=1}^F{W}_{f(mn)}{v}_{f(mn)i}\right)\right].$$ (3) Collaborativeresources’satisfactionaboutdynamicservicemodesBecausedifferentresourceindividualshavedifferentresourcefeatures,theirserviceandcoordinationcostsunderdifferentservicemodesarealsodifferent;theyhaverespectivepreferencesfordifferentservicemodes.Whenanunsuitableservicemodeforthecollaborativeresourceisutilizedtoomuch,thewillingnessofthecollaborationresourcetocooperatewiththeplatformwillbereduced—thisisnotconducivetothelong-termcooperationofbothsides.Therefore,collaborativeresources’preferencestowarddifferentservicemodesneedtobeconsideredwhenconvenienceserviceplatformsadoptingdynamicservicemodescarryoutresourceintegration.Itisthusnecessarytodescribecollaborativeresources’relativesatisfactionunderdifferentservicemodes.Theapproachweutilizedinthisarticleallowscollaborativeresourcestofirstsequencetheirpreferencestowardvariousservicemodesbasedontheirownresourcefeatures.Suppose,fortheserviceitemm,thepreferencesequencevaluesetofthecollaborativeresourceitowardtheservicemodeAmnisRi={\({R}_{in}^1,{R}_{in}^2\),…\({R}_{in}^N\)},inwhich\({R}_{in}^k\)istherankingthecollaborativeresourceigivestothemodeAmn,thesmallerthesequencenumberis,thebiggerthesatisfactionofthecollaborativeresourceitowardtheservicemodeAmn.SupposethesatisfactionofthecollaborativeresourceitowardtheservicemodeAmnisSi(mn),therearemanydescribingpatternsabouttherelationshipbetweenSi(mn)and\({R}_{in}^k\).Thisarticleadopts Si(mn)=1/\({R}_{in}^k\)todescribetheirrelationship,thusSmi,theoverallsatisfactionofthecollaborativeresourceitowarddynamicservicemodes,canbeexpressedas: $${S}_{mi}={\sum}_{n=1}^N\left({u}_{mn}{S}_{i(mn)}\right).$$ (4) ExcavationofkeyfactorsofresourceintegrationunderdynamicservicemodesBeforemakingdecisionsforresourceintegration,itisnecessarytoevaluatealargenumberofoptionalresources,identifythemostsuitableresourcesforintegration,anddeveloplong-termstrategiccooperationwithsuitableresources.Iftakingtheefficiencyofresourceintegrationintoconsideration,platformscannotevaluatealltheaspectsofoptionalresourceswhenanalyzingtheirvalueandevaluatingtheirperformance.Instead,theyshouldcombinethefeaturesoftheplatforms’dynamicservicemodes,evaluateoptionalresourceswithemphasis,clarifythefeaturesneedingthemostattentionwhenconductingresourceintegration,thatis,thekeyfactorsneedingthemostattentionwhenconductingresourceintegration,andcarefullychecktheperformanceofoptionalresourcesinthesecharacteristicaspects,thenusethemasthestandardforsupplierresourceintegration.Itisworthnotingthatwhenplatformschoosedifferentservicemodes,theimpactofthevariousresourcecharacteristicsontheservicelevelwillchange.Forexample,whentheservicemodecontainsspeedservicecontents,thelogisticspeedandinformationinteractioncapabilityoftheservicesupplyresourceswillhaveagreatimpactontheservicelevel,andiftheservicemodeincludescustomizedservicecontents,thecustomizedservicecapabilitiesofthecollaborativeresourceswillhaveagreatimpactontheservicelevel.Basedonthis,Table 3showstheexcavateddominantfactorstowhichmoreattentionneedstobepaidwhenintegratingresourcesunderdynamicservicemodes. Table3ExcavationfordominantfactorsofresourceintegrationofconvenienceserviceplatformsunderdynamicservicemodesFullsizetableVmninTable3indicatesthecorrelationbetweenthecharacteristicsofoptionalresourcesandtheservicelevelunderdifferentservicemodes;wecanassignvaluestoVmn(0 ≤ Vmn ≤ 1)asperactualcorrelationconditions.Combinedwiththeratioofeachservicemodeunderdynamicservicemodes,wecanobtainvariousfactors’overalldegreeofconcerneiduringtheresourceintegrationofconvenienceserviceplatforms.Asmentionedabove,whenintegratingindividualresources,weneedtocarryouttargetedinvestigationsonspecificaspectsofresourceperformance.Therefore,wecansetacertainstandardforei,forexample,setthefactorswitheigreaterthan0.45asthekeyfactorstowhichmoreattentionneedstobepaidwhenintegratingresources,andevaluatetheperformanceofresourcesonthisbasis.ResourceintegrationoptimizationmodelTheoptimizationmodelofresourceintegrationunderdynamicservicemodesisestablishedinthissection.Asplatformscarryingserviceitems,convenienceserviceplatformsneedtoconsidertheinterestsofbothcustomersandservicesuppliers.Hence,duringresourceintegration,weneedtoconsiderthecustomerperceivedvalueofservicelevelononehandandsuppliers’satisfactiontowarddynamicservicemodesontheother.Therefore,thisarticlesetsfourresourceintegrationoptimizationobjectivesunderdynamicservicemodes:maximizationofcustomerperceivedvalueofservicelevel;maximizationofsuppliers’satisfactiontowarddynamicservicemodes;minimizationofresourceintegrationcost;andminimizationofresourceintegrationtime.Toachievethesynthesisofthemultipleoptimizationobjectives,theparametersintheobjectivefunctionsneedtofirstbestandardized.TheparametersandvariablesoftheoptimizationobjectivesaredefinedinTable 4. Table4DescriptionofvariousparametersandvariablesFullsizetableTheoptimizedmodelforresourceintegrationisestablishedasfollows: $$\max{Z}_1=\max{\sum}_{m=1}^M{\sum}_{i=1}^I{E}_{rmi}g\left({r}_{mi}\right),$$ (5) $$\max{Z}_2=\max{\sum}_{m=1}^M{\sum}_{i=1}^I{S}_{rmi}g\left({r}_{mi}\right),$$ (6) $$\min{Z}_3=\min{\sum}_{m=1}^M{\sum}_{i=1}^I{C}_{mi}g\left({r}_{mi}\right),$$ (7) $${minZ}_4=\min{\sum}_{m=1}^M{\sum}_{i=1}^I{T}_{a(mi)}g\left({r}_{mi}\right),$$ (8) subjectto $${L^{\ast}}_{kmi}={L}_{kmi}+{X}_{kmi}{C}_{kmi}.$$ (9) $${C}_{mi}\le{C}_{max}.$$ (10) $$\mid{T}_m-\left({T}_{end(mi)}-{T}_{sta(mi)}\right)\mid\le{T}_{s(m)}.$$ (11) Eqs.(5)–(8)respectivelyrepresentthefouroptimizationobjectivefunctionsforresourceintegration.Eq.(9)istherelationconstraintbetweentheimprovementamountofresource-dominantfactors’operationlevelandthecostinputinresourceintegration.Eq.(10)isthecostconstraintoftheresourceintegrationinput,andensuresthecostofindividualresourceintegrationisnothigherthanthemaximumintegrationcosttheconvenienceserviceplatformscanacceptforindividualresources.Eq.(11)isthetimeconstraintinresourceintegration,andensurestheintegrationtimediscrepancywhenintegratingindividualresourcesisnotgreaterthanthemaximumtimediscrepancytheconvenienceserviceplatformscanaccept.AlgorithmTheoptimizationmodelofresourceintegrationunderdynamicservicemodesisessentiallyaproblemofconstrainedmulti-objectiveoptimizationthatrequiresatrade-offbetweenmultidimensionaloptimizationgoals,therebymakingitanondeterministicpolynomialtime-hardproblem.Theantcolonyalgorithmiswell-recognizedforitsstrengths,likeportabilityandhighglobalconvergencerate,indealingwithconstrainedmulti-targetmodeling(Tangetal.2013;Yao2013;YaoandDeng2015),andiswidelyappliedindynamicmulti-targetoptimizationproblems(Panickeretal.2013;Tangetal.2013;TingandChen2013;Wongetal.2012;Yao2010;Yao2013).Additionally,inthisstudy,thefourobjectivefunctionsareprobablycorrelatedinacomplex,nonlinearorcontradictorymanner,therefore,adynamicweighingmechanismisneeded.Tosolvetheproblem,thealgorithmissupposedtomeettworequirements.First,thedecisionvariableshouldbeofmultipleattributessothatcharacteristicsofdifferentoptimizationtargetscanbesimultaneouslyconsideredintheconvergenceofmultipletargets.Second,thealgorithmshouldbeemployedtocoverallpotentialserviceresourcesinsteadofbeingconfinedtospecificresourcestobetteradapttothecomplexorcontradictiverelationshipsamongoptimizationobjectives.Someavailableliteraturehascomparedantcolonyalgorithmwithothersformulti-objectiveoptimizationproblems,andtheformerseemssuitablegivenitscharacteristics(Dongetal.2007;Merkleetal.2002;Xiaoetal.2011).Thus,theantalgorithmisemployedforthisstudy.Whenconstructingthealgorithm,eachalternativeserviceresourceisconsideredasanindependentunitwithstableoperationalparameters.Supposethattheserviceresourcesnetworkfortheconvenienceserviceplatformsiscomposedofsourcenodes,includingthosetobeintegrated.Theselectionprocessofresourcestointegratecorrespondstotheforagingbehaviorofants.Startingfromthesourcetotheobjectnode,antswillmovethroughtheresourcenodesinthenetworkthendisappear,andthosealternativeresourcesthatdonotmeettheconstraintconditionsofthemodelaresetasforbiddennodes.Thenodesontheforagingpathofantsaretheselectedonesforresourceintegration.Accordingtodifferentoptimizationobjectives,theattractionprobabilityofeachresourcenodetoantsissetasfollows.Settheattractionprobabilityofrmitotheantsaccordingtoruleofthemaximizationofoverallcustomerperceivedvalueofresources’servicelevel.Setπemi,thesortofpheromonequantitiesremainbyantspassingrmi,istheratioofoverallcustomerperceivedvalueofresources’servicelevel.Thus,theattractionprobabilityisgivenasEq.(12): $${P}_e={\varPi}_{emi}/{\sum}_{i=1}^I{\varPi}_{emi}.$$ (12) Settheattractionprobabilityofrmitotheantsaccordingtoruleofthemaximizationofsuppliers’satisfactionwithdynamicservicemodes.Setπsmi,thesortofpheromonequantitiesremainbyantspassingrmi,istheratioofthesuppliers’satisfactionwithdynamicservicemodes.Thus,theattractionprobabilityisgivenasEq.(13): $${P}_s={\varPi}_{smi}/{\sum}_{i=1}^I{\varPi}_{smi}.$$ (13) Settheattractionprobabilityofrmitotheantsaccordingtoruleoftheminimizationofresourceintegrationcost.Setπcmi,thesortofpheromonequantitiesremainbyantspassingrmi,istheinverseratiooftheresourceintegrationcost.Thus,theattractionprobabilityisgivenasEq.(14): $${P}_c={\varPi}_{cmi}/{\sum}_{i=1}^I{\varPi}_{cmi}.$$ (14) Settheattractionprobabilityofrmitotheantsaccordingtoruleoftheminimizationofresourceintegrationtime.Setπtmi,thesortofpheromonequantitiesremainbyantspassingrmi,istheinverseratiooftheresourceintegrationtime.Thus,theattractionprobabilityisgivenasEq.(15): $${P}_t\kern0.5em={\varPi}_{tmi}/{\sum}_{i=1}^I{\varPi}_{tmi}.$$ (15) Accordingtotheaboveanalysis,thecomprehensiveattractionprobabilityofrmifortheantsisgiveninEq.(16),whereλ1,λ2,λ3,λ4(0