evaluatesupiea

D.
Li,Y.
Liu,andY.
Chen(Eds.
):CCTA2010,PartI,IFIPAICT344,pp.
674–680,2011.
IFIPInternationalFederationforInformationProcessing2011EvaluationontheAgriculturalWebsite'sEfficiencyBasedonDEAMethodShangminDeng1andWeiliMen21Headmaster'sOfficeofShandongUniversityofTechnology,Zibo,P.
R.
Chinadeng2132@sdut.
edu.
cn2ScienceandTechnologyInformationResearchInstituteofShandongUniversityofTechnology346380987@163.
comAbstract.
Aninformation-drivenagricultureisfirstlyagriculturedigitalization,thefoundationofwhichistheconstructionofagriculturalinformationnet-works.
Basedonthemethodsofwebsiteevaluation,anindicatorsystemofeffi-ciencyevaluationforagriculturalinformationwebsitesisconstructedinthisthesis.
Thentheefficiencyof17municipalwebsitesinShandongprovinceareanalyzedbytheC2RModelinDEA.
Theresultdeclaresthat3oftheseweb-sitesarerelatedeffectiveandtheother14arenon-effective.
Byananalysisofprojection,thisthesisalsogivessuggestionstoimprovetheperformanceofthenon-effective.
Keywords:Agriculturalinformationwebsite,Efficiencyofwebsite,DataEn-velopmentAnalysis(DEA),Evaluationresearch.
1IntroductionIssuesconcerningagriculture,countrysideandfarmers,arecalledThreeRuralQues-tions,whichtakeancrucialpartintherealizationofthegoalofawell-to-dosocietyandthedevelopmentofourcountry'seconomy.
Itisacknowledgedthatitwillbeharderforuninformeddistrictstodeveloptheirownagriculture.
Nowadays,theinformationgapbetweenurbanandruralareashasattractedwideattentionofexpertsfromInformationScience.
ThePartyCentralCommittee,since16thpartycongress,hammersawayatthepointthatweshouldalwaystakeaddressingtheproblemsfacingChina'sagricultureasthetoppriorityofoureconomic,andalsoputforward"toboostthecountyeconomy"atthesametime.
Oneof"theThreeMostRequire-ment",raisedontheThirdPlenarySessionofthe11thCentralCommitteeoftheC.
P.
C,isthatthepositionofagricultureasthefoundationisstillweakandneedtobrushuponthemost.
InMarch2010,thegovernmentofShandongprovincemakestheimplementationoftheplannamed"PowerfulProvinceofinformation"andthedevelopmentofDigitalAgriculturepositivelyaskeypointstopromotenewInforma-tionIndustries.
[1]EvaluationontheAgriculturalWebsite'sEfficiencyBasedonDEAMethod675Theinformationizationofagriculture,inthefirstplace,isdigitalization,whichmeanstodevotemajoreffortstodevelopthedigitalagriculture.
Asweallknowthat,thefoundationofdigitalizationistheconstructionofinformationnetworksandtheservicelevelisdeterminedbysuchfactorsastheresources,theinfrastructuralfacili-tiesandtheenvironmentofinformationization,aswellasthehumanresourcesandtheeffectivenessofagriculturalinformationservices.
Inanotherword,theinforma-tionizationofagriculturalservicecanbedemonstratedbythatofagriculture.
[2]So,itisreasonabletoestimatethepresentsituationofagricultureservices.
However,be-causeofthedifferenceingeographicalconditionsandculturebackgrounds,weneedaappropriatemethodtoselectedsamplescomprehensivelyandobjectively.
FortunatelythereisasuccessfulprecedentcreatedbytheDEA(DataEnvelopmentAnalysisforshort)whichisconvenientandpractical.
[3]2DataEnvelopmentAnalysisDEAwhichisshortforDataEnvelopmentAnalysisisanewsystemanalyticme-thod.
Itwasfirstlyputforwardbytwooperationalresearchexperts,A.
CharnesandW.
W.
Copper,todealwithinputsandoutputssubjecttorandomdisturbances,accord-ingtotherelativeefficiencyandevaluationofreturnsforfamiliardepartments.
Basedontheconceptofrelativeefficiencies,thismethodevaluatestherelativeefficiencyofDMUs(DecisionMakingUnitsforshort)bylinearprogrammingmodel.
BesidesevaluateandlistDMUsasrelativeefficiency,theresultsalsowillpointoutexistingproblemsandimprovementfornon-efficientDMUs.
Furthermore,insteadofsettingtheweightofinputs/outputsbeforehand,DEAgetsobjectivevalueforeachrandomvariablebycomputingactualdatas,whichmakestheconclusionindependentlyandimpartially.
[4]Inthelightofdifferentresearchpurposes,therearemanydifferentDEAmodels.
Inthisthesis,weselecttheC2RmodeljustbecausewewanttostudywhethertheDMUsaretechniqueandscalesimultaneouslyefficientwhentodecreaseitsinputsbysuit-abletimeswithitsunchangedoutputs.
TheDMUistechniqueorscaleefficientunlessitisDEAefficient;otherwise,itwillbenotefficient.
3DescriptionoftheModelofC2RAsthefirstevaluationmodel,theresultsofC2Rcanbeusedtodemonstratetheeffi-cientofDMUs.
ForthesakeofourdiscussionweassumethatthenumberofDMUsisn,andthenumberofinputandoutputvariableseachDUMhasis"m"and"s"respec-tively.
wealsosetupamathematicalformulaoftherelationshipforthetwokindsofvariablestofulfill:Xj=(X1j,X2j,X3j,…,Xmj)T≥0,Yj=(Y1j,Y2j,Y3j,…,Ysj)T≥0,j=1,…,n;Atthesametime,othercertainconditionsshouldbesatisfied:Xij>0,Yrj>0,i=1,2,…,m;r=1,2,…,s.
676S.
DengandW.
MenThemodelisknownasefficiencyevaluationofdecision-makingunitssuchasj0,subjecttomeetingtheefficiencyofalltheDUMsatcheckpoints,thenwecometoaC2Rmodlewithnonarchimedeaninfinitesimalvector.
TheoptimalityisrelativeefficiencyofaDMU,ascontrastedwithotherDUMs,accordingtowhathavewesuggested.
[5]Here,weuseθastheinputproportionalvariable,εasthenonarchimedeaninfini-tesimalvector(usually,weletitbe10-6);=(1,1…,1)∈Em,=(1,1…,1)∈Es;isslackvariablescorrespondtoinputs,=(…)isthevariabletooutputandtheλjisthedecisionvariablefortheappointedDMU.
Whenthismodelisappliedtodata,wewillgettheoptimalsolutioncomposedofλj,S*-,S*+andθ*.
Thereareseveralresultsthatweshouldmakeintoourconsid-eration:(1)Whenθ0=1,s0-=0ands0+=0,wethinktheDMUisefficiencywhichmeanstechniqueandscalearesimultaneouslyefficient.
(2)Whenθ0=1,-,buts0-≠0ors0+≠0,webelievethatundersomeconditionstheDMUisweakDEAefficient.
Inotherwords,theeithertechniqueorscaleisefficient.
(3)Whenθ0upieaobjectively,weneedtoclassifieddiffer-entupdatingperiodsinsorts.
Inthisthesis,wedividetheperiodsintofourcatalogues.
Accordingtothedifferentinfluence,wegiveweightstotheperiodsrespectively.
NBLmeansusingthiskindoflinks,wecannotfindthetargetwebsite.
Thoughthis,wehaveanotherwaytomeasuretheeffectiveofwebsites.
ThemoreFRPis,thelong-ertimeitwilltaketooperatethewebsite.
NRPhavethesimilarfunctionasCP.
Bothofthemevaluatetheeffectivenessofwebsitefromtheeyeofsitenavigations.
Output(outpforshort)variableswechoosemeansthenumberofvisitorstothesite,whichhasacertainrelationwithTFS.
Becauseavisitormayglanceovermanydifferentdocumentsormorethanonewebsites,theclicksmaybemuchlarger.
[9]4.
2SampleTodeterminetheAgriculturalWebsite'sEfficiency,wechoose17sitesasDMUsset,builtbyAgriculturalInformationCentreofShandongprovinceandsubordinatede-partments.
Theyhavecommoninmanyaspectssuchascontents,receiversandfunc-tions,whichmeettherequirementsforapplicationofC2RmodelinDEA.
4.
3DataUsingMaxamine[10],aspecializedprofilertomakeaanalysisofnetworks,themen-tionedsiteswerescanned.
Theobjectivedataswecollectedincludestheinternalin-formationofthesitesandinterconnectioninformationbetweennetworks,amongwhich,theUDRarecalculatedastable1.
Theoutputconsultverifiedtestresultsfromalexa.
com,anauthoritativeThird-partycertificationwebsite,applyingthenumberofvisitorforthetargetedsite.
Table1.
CalculationmethodsofupdaterateofwebsiteupdatecycleofpagesupdaterateRiWeightλupdaterateofwebsite≤1weekR10.
501week~6weeksR20.
306weeks~6monthsR30.
15≥6monthsR40.
05678S.
DengandW.
MenTable2.
AppraisalindexoftheefficiencyofwebsitesinshandongprovincenameofcountiestheefficiencyofwebsitenameofcountiestheefficiencyofwebsiteQingdao1.
000Dongying0.
284Weihai1.
000Weifang0.
243Jinan1.
000Binzhou0.
143Zibo0.
662Linyi0.
122Liaocheng0.
494Dezhou0.
057Laiwu0.
439Tai-an0.
055Jining0.
395Zaozhuang0.
050Yantai0.
362Rizhao0.
042Heze0.
315Formtable2,wecanseethatthreeDUMSareofeffectiveunitsinDEA.
Inotherwords,theothers14arenonDEAefficient,amongwhich,therearefourwebsiteshaveaneffectivenessbelow0.
01.
Therefore,itcanbeconcludedthat,theagriculturalsites,efficiencyinShandongarelowerthanexpected.
Then,wemakeanprojectionanalysisforthenonDEAefficientbasedontheDEAP2.
1results.
Themaintaskistoanalyzerelativeimprovementofinputswhenoutputsisremainunchanged.
Wealsocalculatepotentialimprovementsofinputsandoutputsunderidealcondition.
Theaveragepotentialimprovementsareshowedintable3.
Thepotentialimprove-mentsdemonstratethat,referencetothenonDEAefficientDMUs,wehavenotmakefulluseofeachinputalready,so,intheoryatleast,wecanpromotethesiteseffi-ciencyandinduceinputs.
Ifwekeepinputsstill,theoutputisalsocanbepromotedby61%.
Inotherwords,ifinputscannotbereduced,theoretical,efficiencycanberaisedbytheincreasingnumberofvisitors.
Table4showsmainfactorsinfluencingwebsitesefficiency,includingFRP,TAG,NRPandAPS.
4.
4RecommendationsFromtable2,weknowthattheoperationefficienciesareverydifferentamongthese17municipalagriculturewebsites,eventhoughmostofthemareinefficient.
Refer-encetononDEAefficient,combinedwiththeanalysisoftable3andtable4,wesuggestthattherelativedepartmentsshouldtakesuchactionsasfollowingtoimprovetheperformanceofthenonDEAefficient:Table3.
ThemeanpotentialimprovementsforinputandoutputindexesLWHP,QS,QS,QS,QS,QS,QS,QS,QS2XWS7$*$367)68'51%/)35153&35HDFKPHDQEvaluationontheAgriculturalWebsite'sEfficiencyBasedonDEAMethod679Table4.
AnalysisonthemaininfluencefactorsforinefficientDMUswebsiteComparedItemwebsiteComparedItemwww.
zbny.
gov.
cnAPSwww.
wfny.
gov.
cnTFS,FPRwww.
lcagr.
cnTAG,FPR,NRPwww.
bzny.
gov.
cnTFS,NRPagri.
laiwu.
gov.
cnTAG,APS,NBL,www.
lyny.
gov.
cnTAGwww.
kelawang.
comTAG,FPRwww.
dzny.
gov.
cnTAG,NBL,NRPwww.
ytny.
gov.
cnTAG,UDR,NBL,www.
tany.
gov.
cnTFS,FPR,NRPwww.
hzsny.
gov.
cnTAG,APS,UDR,www.
zznjh.
gov.
cnFPR,NRPwww.
dyny.
gov.
cnFPRwww.
rzny.
gov.
cnAPS,FPR,NRPAbstractcontents.
ThetargetofthisactionistoreduceTAGandlowerAPS.
Thetwoaspectsnotonlyreflecttheinformationwebsitecontents,butalsothewebsite'susabilityandsecurity.
Agriculturesitesaremadeforagricultureparticipantsandrelativedepartments.
Soearlyinthewebsiteconstruction,weneedtoconsiderhowtodesignthepagesreasonably,andabstractthecontenttoensureitsusabilityandsecurity.
CutdownNRP.
NRPevaluateswebsitesfromtheperspectiveofnavigation.
Thisaspectmeansreducinginputsofnavigation.
IfthesitehasasetRobustness,themorecomplexitythenavigationis,theeasiervisitorswillbelost.
This,finally,willindi-rectlyaffectthewebsite'sefficiency.
Ifthenavigationissimplified,notonlytheus-abilityisimproved,butalsotheinputsoftheoperationcanbecutdown.
ReduceTFS.
Thismethodincludesreducinginputsofwebsitepromotionorincreasethenumberofthevisitors.
Onsomedegree,TFSrepresentsthestabilityofawebsite,yetthemoreexternallinksare,thehigheritwillcost.
Checkingdeadlinks,updatingoldlinksandotherresourcesmaintenancewillaffecttheoperationefficiency.
So,inordertopromotetheefficiency,weneedtoreduceTFSwhilethequalityofthewebisensured.
Onlyinthisway,canweattractmoreattentionandkeepthesiteauthority.
5ConclusionsThoughalltheabovementioned,wehaveaclearideathatthereare14websitesarenonDEAefficient,whichcountsfor84.
3%ofallmunicipalwebsitesinShandongprovince.
Wehavepresentedadetailedaccountingofefficiencyofagriculturalweb-sitesandsuggestionsarealsogiventoimprovethecurrentsituation.
Evaluationoftheagriculturewebsites'efficiencycannotonlyreflectthepresentsituationofShandongagriculture,butalsobevaluableforfurtherdevelopmentintheconstructionofinfor-mationagricultureandapplicationofrelatednetresources.
680S.
DengandW.
MenReferences1.
Liu,S.
H.
:StudyontheIndicatorSystemforMeasuringtheRuralAreaInformatizationLevelinChina.
J.
LibraryandInformationService,33–36(1981)2.
Han,X.
S.
,Pan,H.
F.
,Wen,J.
H.
,etal.
:EvaluationandStudyonDevelopmentofAgricul-tureInformationService.
J.
JournalofAgriculturalMechanizationResearch,20–23(2007)3.
Hu,P.
,Fu,Y.
Y.
,Gan,L.
:EvaluationonagriculturesitesinwesternareaBasedonDEAandClusteranalysis.
J.
ScienceandManagement,20–22(2005)4.
Du,D.
,Pang,Q.
H.
:ContemporaryComprehensiveevaluationmethodsandclassicdocu-mentation.
TsinghuaUniversityPress,Beijing(2005)5.
Lan,F.
:TheEvaluationofChinasCirculationEconomyEfficiencyBasedonDEA.
J.
JournalofHarbinUniversityofCommerce(SocialScienceEdition),16–19(2000)6.
Wu,L.
:Evaluationontheefficiencyofeco-technologyInnovationbasedonDEA.
J.
TechnologyProgressandPolicy,114–117(2009)7.
Tan,L.
,Wang,X.
C.
:StudyontheEvaluationofAgricultureWebSite.
JournalofAnhuiAgriculturalSciences,1876–1882(2007)8.
Sun,Y.
L.
,He,Y.
,Zhao,Z.
N.
:EvaluationontheWesternAgriculturalWebsitesEffi-ciencyBasedonDEAMethod.
JournalofIntelligence,14–17(2009)9.
Qiu,H.
:TheoryResearchandApplicationforDEAinwebsitesevaluation.
CambridgeUniversityPress,NewYork(2005)10.
WebAnalystSoftwarePackage,http://www.
maxamine.
com