2016美賽原版獎(jiǎng)-優(yōu)秀已打2016d

Forofficeuseonly

D

Forofficeuseonly(Yourteam'ssummaryshouldbeincludedas pageofyourelectronicysisofSociety’sInformationWiththedevelopmentoftechnology,thesituationofinformationflow esmorecomplicatedwhichhasarousedwidelyconcerns.Inthispr,weestablishadouble-layercomplexnetworkmodeltomeasuretheevolutionandinfluenceinsociety’sinformationnetworks.Inourdouble-layeredcomplexnetworkmodel,theinnerlayeristoyzethesituationoftheinformationflowamongaregion.Wespeciallytakeseveralfactorssuchasinherentvalueofinformation,themediaeffectand alsubjectiveemotionsintoaccounttofytheattributesofnodeandside.OptimizingtheAsICModel,weestablishpointtopointmodeltosimulatetheflowofinformationinthenetwork.Intheouterlayer,BasedontheReaction-DiffusionModel,weestablishaglobalnetworkmodeltocorrecttheformulasofpropagationprobabilityandpropagationdelaybytakingthedistancefactorintoconsideration.Basedonourconstructedmodel,withthecombinationoftheinherentvalueoftheinformationandtheefficiencyofinformationflow,wegiveoutthemethodofqualifyingnews.Accordingtothedifferentconditionofmediausage,weobtaintheproportionofdifferentmediaindifferentareasbasedonregressionysis.Comparedthesimulationofpropagationratioinournetworkwiththeactualmarketshareofacertainmedia,wefindthattherelativeerrorislessthan10%.Basedonourcollectingdata,wepredictthatthepropagationprobabilitywillincreaserapidlywhilethepropagationdelaywilldecreasearoundtheyear2050.Meanwhile,thesignificanceofinherentvalueofinformationwilldepress.Later,basedonBoundedConfidenceModel,weestablishaninteractionmodeltoyzefactorswhichinfluencethepublicinterestandopinionthroughinformationnetworks.Asaresult,wegetdistributionfunctionofpublicopiniontoobservethechangeofpublicopinion.Throughapieceofspecificinformation,wefindthepublicopinionchangesrapidlyinthebeginningandfinallyintothestablestate.Next,weyzetheinfluenceofseveralfactorsinspreadinginformationandpublicopinion.Wegetdiagramaftercalculationtoshowthetrend.Forexample,thepropagationprobabilityandchangerateofopinionareproportionaltoinformationvaluewhilepropagationdelayisopposite.Meanwhile,weadoptsensitivityysisondifferentelements.WeconcludethatthemodelismoresensitivetoAuthority,latercomesInfluence,ActivityandWilling.Keywords:double-layercomplexnetworkmodel,pointtopointmodel,BoundedConfidenceModel,information Our Task1:EstablishaDouble-layerComplex InnerLayer:InformationFlowovera Index Informationflowbetweentwo OuterLayer:InformationFlowfromRegionto ChangetheValueof Removea Task3:Predictthesituationaroundtheyear Task4:InfluenceinPublicInterestand Foundationof InfluenceinSpreading InfluenceinPublic Strengthsand Nowadays,informationisspreadquicklyinourdailylife.Duetothedevelopmentoftechnology,nomatterthebigeventsorthetrivia,peoplehaveaccesstoinformationquicklyandconveniently.Asaresult,theinformationgarbageiseverywherewhichinfluencesourlife.Inordertomanageandtracktheflowofinformation,peopleraiseawarenessoftheimportanceofestablishingasocietyinformationnetwork.Interestedininformationcommunicationsituationofsocialnetwork,severalmethodshavebeenadoptedtodescribeit.Mostmodelssupposedthatinformationisspreadaccordingtotimesequence.Forexample,IndependentCascadeModel[1]regardsthatpropagationprobabilityissameamongnodesallthetimewhichhasnomemory.Meanwhile,Linearthresholdmodel[2]concentratesoncumulativeprobabilitywhichreflectstheacceptedabilityofreceiver.However,thesemodelsignoresuchfactors:Attributesofnodessuch alInfluenceofinherentvalueofinformationinPropagationprocessisnotTherefore,thereisanurgentneedforacompletescientificmodel.Basedonit,weestablishadouble-layercomplexnetwork.OurTofurtherpresentoursolutions,wearrangeour rasInsection2,wegiveoutthereliableassumptionstosimplifytheInsection3,thedouble-layernetworkisconstructedtoyzeandforecasttheinformationflow.Theinnerlayeryzestheinformationflowamongaregion.Whiletheouterlayeryzetheflowfromregiontoregion.Meanwhile,thesensitivityysisisgivenouttovalidateourInsection4,weapplyourmodeltosolveaseriesofproblem,suchasqualificationofnews,predictionofinformationflowtodayandaroundtheyear2050.Next,weestablishamodeltoyzethefactorstoinfluencethepublicinterestandopinion.Later,weyzetheinfluenceofdifferentindexesinSpreadingInformationandPublicAtlast,wediscussthestrengthsandweaknessesofourmodelinThedatawefoundisauthenticandNoinventionofmediainthetimeforprediction.Thatistosay,peopleusethesame today,justtheproportionofmediachanges.Ignorethefirewallamongcountrieswhichrestrainsthespreadofinformation.邁思數(shù)模2018美賽課程表及報(bào)名安排事項(xiàng) VIP學(xué)員無(wú)報(bào)名費(fèi)（強(qiáng)制80（建議參加）；點(diǎn)評(píng)輔助報(bào)名：750元模擬賽學(xué)員：700門(mén)班課程元；三科398元【限時(shí)特惠168-268-368】賽前沖刺20節(jié)課法班課程賽前沖刺課程更新時(shí)間為：11共計(jì)20節(jié)課賽前沖刺20VIP（入門(mén)班課程【限時(shí)特568-668-768擬賽+輔助報(bào)名（100，支付后，請(qǐng)聯(lián)系下VIP（算法班課程擬賽+輔助報(bào)名（100擬賽+輔助報(bào)名（100進(jìn)群（指導(dǎo)將通過(guò)提醒：VIP班學(xué)員強(qiáng)制參加模擬賽，其他學(xué)員建議參與模擬賽，因?yàn)楦鶕?jù)國(guó)賽模擬賽結(jié)果來(lái)看，參與模擬賽的學(xué)員幾乎都了國(guó)獎(jiǎng)，未參加模擬賽的學(xué)員國(guó)獎(jiǎng)的比例遠(yuǎn)沒(méi)有參加模擬賽國(guó)獎(jiǎng)的比例高?。?！聯(lián) ；informationintheournetworkisnotDuetothedifferenceofdevelopmentlevelisrelativelylittleinasmallarea,weignorethedifferenceinmediausagefromastatisticalpointofview.Tosimplifythemodel,weonlyconsidersixmediatospreadinformationwhichare r,egraph,radio,evision,theInternetandmobilephone.Thenodespreadthesamemessageonlyonce.Thatistosay，thein-degreeofanodeisnomorethan1.Task1:EstablishaDouble-layerComplexAlargeamountoffactorsoughttotakeintoaccounttofytheflowofinformation.Forexample,duetothedifferenceindevelopmentlevelamongcountries,thewayandspeedofflowmustbedifferent.Also,withtheInternettechnology emoresophisticated,eachplaysamoreandmoreimportantroleinspreadinginformationinsteadoforiginalmediasuchasnewsp rand evision,whichtheobjectivefactorssuchas alinterestsshouldbetakenintoconsideration.Basedonaboveysis,wecanseethatasimplenetworkcannotreflectthewholesituationoftheflowofinformationwell.Asaresult,acomplexnetworkwithmulti-layers,whoselayershaveacertaininteractionsandindependence,seemstobemorereasonable.Therefore,wecomeupwithadouble-layerednetwork,globalinformationnetworkandpointtopointinformationnetwork,whichbelongstoacomplexnetwork,toreflecttheinformationflowmorespecifically.Thestructureofourdouble-layercomplexnetworkisshowninFigureFigure1Adouble-layerThebluedotsrepresenteach andsmallcirclerepresentsaregion.Thedirectedlinerepresentsthedirectionofinformationflow.Thefigurevividlyindicatesthattheinformationoriginatesfromanode(source),thenspreadquicklyoveraregionandatlastspreadthewholeworld.Theinternalandexternalfactorswhichinfluencetheinformationflowisyzedindetailinnexttwosections.InnerLayer:InformationFlowoveraTheprocessofinformationpropagationcanbedividedintomassivepointtopointcases.Eachcasehasthreeentities:sender(S),receiver(R)andcontent(C).Figure2showstheFigure2RelationshipsofthreeentitiesinthediffusionTheinformationspreadfromsendertoreceiver,eachspreadcasecarriesapieceofIndexInordertodescribetheflowofinformationspecifically,wechoose asanexample.BasedontheAsICModel,toimprovetheaccuracyofourmodel,wetaketheinherentvalueofinformation,themediaeffectand alsubjectiveemotionsintoconsiderationandgivefiveindexestoyzetheinformationflow:CommunicationMediaEffect,SenderCharacteristic,ReceiverCharacteristic,ContentCharacteristic,RelationbetweenSenderandReceiver.Supposethatnodeuasasender,nodevasareceiver.CommunicationMediaCharacteristic(mFromahistorical,thecommunicationmediachangealot.Fromnewspr,andradiointheearlytimetoevision,Internet,mobilephoneandmanyothermedianowadays.Informationtendstospreadmorequicklyandwide.Duetothedifferenceinsocialdevelopments,mainstreammediaofdifferentregionsvariesalot.Basedonsixmedia(newspr,egraph,radio,evision,Internetandmobilephone),communicationmediumeffectmiscalculatedasmii(iN ,R,TV,I,Mi

WhenN,,R,TV,Iisshortforsixmedia,newspr,egraph,radio,evisiontheInternetandmobilephonerespectively;whereidenotesthemarketshareofithmediumandidenotescorrespondingeffectonspreadingcommunicationsuchaspropagationspeedandscope.BaseontheyticHierarchyProcess[3],wetrainthecollecteddataandobtainthefinalweightsasfollows.Theresulthasbeenapproximatedinordertosimplifycalculation.Meanwhile,inaspecificareaweassumethatiandmisconstantwhichrepresentaveragelevel.TheweightofdifferentmediaislistedinTableTable1Differentmediaandtheireffectonspreading132MobileSenderCharacteristic(sInfluence(F):Influenceshowstheimportancedegreeofanode.ItshouldInF(u)lg(RTWhereRTdenotesthetotalnumberofforwardingtimesofa

Authority(A):Itisdefinedastheratioofin-degreedividesout-degreewhichiscalculatedasfollows:A(u)lg(FOWhereFOdenotesthenumberoffanswhileFRdenotesthenumberof

Ac(u)lg(POD

WherePOdenotestotalnumberofmessagestheuserhasbeenpublished.Ddenotesthenumberoftheuser’sactivedays.Wetakenormalizedformmembershipfunctionsforeachindexsothatvaluesofallthefactorscanbeconstrainedbetween0and1.Thenormalizationisshownasfollows:ximin{xjx' 1

max{x}min{x1 1 Wherexandx'respectivelydenotestheoriginalvalueandthevalueafter ReceiverCharacteristic(RWilling(W):Thevalueofwillingindicatesthepossibletrendforreceiverstospr-eadtheinformation(v)iscalculatedasfollows:W(v)lg(RPWhereRdenotesthenumberofretweettimes.PdenotesthenumberoforiginaliInthesameway,normalizetheindexasy,i

ContentCharacteristic(cInherentvalueofcontentplaysanimportantroleintheinformationflow.Forexample,thefunnierormoreimportantmessagetendstospreadmorequickly.Toevaluatesuchidea,wefyitfortwoaspects:thevalueofinformationcontentandthequalityofLuckily,thearticleinJournalofInformationEngineeringUniversity[4]determines5indicatorsforthevalueofinformationcontentand5indicatorsforthequalityofpresentation.SpecificindexesandtheirweightsareshowninTable2.Table2Evaluationindexesandtheir InherentQualityofClearCarefulConciseComprehensiveWegiveoutgeneralformulastocalculatetheabovetwoaspects 5

10 value 10

i

c0.7valueidenotestheweightofabove10indicatorsrespectively.

denotestheofthecorrespondingindex,whicharrangesfrom0to9.valuedenotesthevalueofinformationcontentandpredenotesthequalityofpresentation.PQPPQPSi(u,v) wherePandQrespectivelyfilevectorofusesuandStructuralsimilarity(Ss):Itillustratesthesimilarityofcircleoffriendsamongusers.JaccardDistance[6]isappliedtotheSs(u,v)

N(u)NN(u)NN(u)NN(u)NCloseness(CL(u,v)):Whenthemessagecontainsinformationofreceiver,itwillbespreadmoreeasily.DefineCL(u,v)asfollows:0,ontheCL(u,v)1,whenthemessage0,ontheInformationflowbetweentwoBasedonabove ysis,weestablisheigenvectorofnodevector(n)andsidevector(e).Thesetwovectorsareshownasfollows:n

CeS CToobtainthepropagationprobability,fundamentallyfunctioniscalculatedasf(u,v,c)kT 1 2Wheremisaconstant,1denotesweightofnodes’characteristic,2denotesweightofsides’characteristic.Theweightrepresentshowcorrespondingcharacteristicinfluenceinpropagationprobability.k1istheweight.BasedonBayesianlogisticfunction[7],thepropagationprobabilityisp(u,v,c) 1In(f(u,v,Inthesameway,propagationdelayiscalculatedas

(u,v,c)kT 1 2Wheremisconstant, 1and2respectivelydenotesweightofnodesandsidescharacteristics.k2istheweight.Inordertocalculatethetheory[8]isadoptedas

ki,iandi umlikelihoodTakingtimedecayfactorintoaccount,thepropagationprobabilitydecreasesastimegoesby.Accumulatedpropagationprobabilityfromnodeutonodevisdefinedasfollows:uF((v,tv)|(u,tu))tf(u,v, u

ti(iu,v)denotesthetimewhennodeispreadtheS((v,tv)|(u,tu))astheprobabilityofwhichnodeuwillnotspreadinformationtovbeforetimetvS((v,tv)|(u,tu))1-F((v,tv)|(u,tu DetermineasetDk{(v,t),...,(v,t)}torepresentsallthenodeanditsi ntime.AndthesetQk{(v,t)Dk,tt}denotesthenodeswhichspreadi beforetimet.par(v)isdefinedastheinformationsourceofnodev,thatistosay,theinformationspreadfrom par(v)tov.Thepropagationofapieceofinformationckiscalculatedasf(ck)F((v,tv)|(u,tu)) S((v,tv)|(u,tu

(v,tv

Qk(v)\Sothepropagationprobabilityofallpiecesofinformationinsetbeobtainedinthefollowingformula:

C Togetthelargervalue

f(C)f(ck1kf(C),wesettheobject

minlgf Traindatathroughcomputer,wecanfinallyobtainthevalueofiOuterLayer:InformationFlowfromRegionto

ki,iAsyzedabove,thedevelopmentlevelisunbalancedontheEarthwhichinfluencestheflowofinformation,sothenodeswechoosemustberepresentative.Tosimplifythemodel,wechooseGDPpercapitaasthecriteria,withdatafromtheCJDBY.net[9],whichisshowninFigure3.Figure3GDPper Figure4GlobalWecanseethateachofwhichhasobviousdifferenceGDPpercapitaissignedbyacertaincolor.Apparently,theGDPpercapitaofAmerica,Canada,NorthernEuropeandAustralia(areawithcolorofpurple)isfourtimesasbigastheworld’saveragelevel.Onthecontrary,theoneofareawithdarkbluecolorisrelativelow.Accordingtodevelopmentlevelandgeography,wechooseseventypicalcountriesinFigure4:America,Brazil,France,Congo,Russia,ChinaandAustria.Inordertoinvestigatetheflowofinformationamongnodes,wedetailthisphenomenonbyapplyingthemechanismoftwo-dimensionalreaction-diffusionmodel(R-Dequation).Ascalarfieldf(,)anditsgradient,whosemagnitudeis umrateofoff perunitlengthofthecoordinatespaceatthegivenpoint,couldbedescribedbelowinpolarcoordinatesystem: V[f(,)]fe1

Therefore,thisscalarfieldf(,)willtransferintoavectorfieldVaftergradient.Forasourceflow,allthestreamlinesarestraightlinesemanatingfromacentralpoint,asshowninFigure5.Obviously,weseethatthecomponentsintheradialtangentialdirectionsaref/andf/,respectively,wheref/0Figure5ApointsourceAccordingtothemechanismofreaction-diffusionmodel,thedivergenceofeverypointinthisvectorfieldVcanbeobtainedeasily:divdiv(f)div(fe 1(f)1f2 Where:div()denotesdivergencewhile()denotesThesecondpartinaboveequation,thatis,2f/2,haslittleimpactsonthefinalvalue.Here,fortheconvenienceofcalculations,weoverlookthispart.di(vf)

AndfinalconclusioncouldbedrawthatdivergenceofgradientofascalarfieldinverselyproportionaltotheradialConsideringthefactthatthefurtherdistanceleadstoweakerpropagationofinformation,soweonlytakethelinkbetweenanodeanditsneighboringnodeintoaccount.Figure4showsthenodeschosenanditsdirectconnect.Theyellowlinerepresentsinternationalinformationflow.Fromaboveysis,weassumethatCommunicationMediaEffect(m)isaconstantamongacountrywhichignoresthedifferenceofdevelopmentlevelamongtherangeofcountries.Meanwhile,thefactorofdistanceisneglectedinpointtopointinformationnetworkaswell.Tofullpresentthereality,basedonabovemodel,wetakenintodistancebetweentwocountriesanddifferenceofmintoaccount.Itisacommonsensethatpropagationinformationbetweenpointsisinverselyproportionaltotheirdistancewhilethepropagationdelayisonthecontrary.Andtheinformationflowislimitedtorelativelybackwardcountry.Comparethembetweentwocountriesandchoosethelowervalueofmintocalculation.Accordingtoaboveassumption,togetherwiththemechanismofreaction-diffsionmodel[10],wesupposethatthepropagationofinformationhappensfromnodeutonodev.Thepropagationprobabilitypexandpropagationdelayexbetweenuandvcanbedescribedas:

(u,v,c)

Where:duvdenotesthedistancefromnodeutonodev,andk1,k2denotestheweightrespectively.InformationFlowWehaveobtainedthepropagationprobabilityandpropagationdelayfromaboveysis.Setuptwosets:setKincludesthenodeswhichhavespreadtheinformation(includingtheinformationsource)andsetNincludesallnodesexceptnodesinsetK.Toexpresstheinformationflowclearly,therearemainlyfourstepsasStep1:Supposethenode1astheinformationsourceandthenumberofallnodesm.K1.10(idenotesthespreadtimefromsourcetonodei).n1(ndenotesthenumberofnodesinsetK).Step2:Calculatethepropagationprobabilityf(i,j)andpropagationdelaybetweenthenodeiinsetKandalltheirneighboringnodejinsetN.

(i,jStep3:Iff(i,j)(isathreshold,usuallyfrom0.5to0.6),takenodejintosetnn1,ji(i,j).ReturntoStepStep4:Iftherenotexistsanodevwhichf(u,v)andthelastincludednodep.Theconclusionisthatpisthetotalpropagationtime,n/misthetotalpropagationratio.Figure5and6helpvisualizespreadofFigure5Originalnetwork Figure6InformationspreadingnetworkThebluespotsrepresentsthenodeswhichhavespreadtheinformation.AsshownintheFigure5,thereisonlyonenodeinthebeginningwhichisinformationsource.Throughthespreadofinformation,asshowninFigure6,thenumberofbluespotsincreasesandnewbluespotsworkasnew‘informationsource’tospreadSensitivityChangetheValueofBasedonourconstructedmodel,weextractmanyindexesastheattributesofnodeandside.Forthepurposeofconcisewriting,wechoosetworepresentativeindexesfordetaileddescriptionwhichrepresentstheattributesofnodeandsiderespectively.Figure7SensitivityysisofW Figure8SensitivityysisofSiFigure7andFigure8showthesensitivityysisofwillingdegreeofreceiver(attributesofnode)andtheinterestsimilaritybetweenthesenderandreceiver(attributesofside).Thebluesolidlinerepresentstheoriginaltrendofwillingdegree,thedashedlinesrepresentadjustedtrendsofindex;redonematchesexpandingvalueofindex,whilegreenonematchesreducinginfluenceofindex.ItisobviousthatthetotaltrendofthreelinesinFigure13isallrising,whichrepresentsthatthepropagationratioisinproportiontowillingdegree.Increasingthevalueofwillingdegreeseemstohelpimprovetheflowofinformation.TheresultsofsensitivityysisofSiissimilartotheoneofW.ButthetrendforSiissmootherwhichshowsthatourmodelisnositivetotheinterestsimilaritycomparedtowillingdegreeofRemoveaToshowthedifferentinfluencedegreeofdifferentindexestotheflowofinformation,wechoosefourindexes:Influence(F),Authority(A),Activity(Ac),Willing(W)asFigure9showsthedifferentpropagationdelay(whenthepropagationratioreaches10%)ofthecasewhenweremoveoneofabovefourindexes.Figure9SensitivityysisofRemovingaWecanconcludethatthelackofInfluenceindexandWillingindexleadstoincreasingthepropagationdelaytime.ThedifferencebetweenthecaseoflackFandthenormalislargerthanthecaseoflackWandthenormal,whichrepresentsthatFismoresensitivetoourmodelthanW.Inthesameway,thelackofAuthorityleadstodecreasingthepropagationdelaytime.Amongallthesefourindexes,modelismostsensitivetoauthority,latercomesinfluence,ActivityandWilling.Meanwhile,thedifferenceisrelativelysmallwhichindicatesthestabilityofourApplication Afterallequationsaredeterminedabove,ourmodelisfinalcompleted.Sonext,weconcentrateonsolvingrelativeproblemsandgivingoutysisaswell.Task1:QualificationofBoththeinherentvalueoftheinformationitselfandspeedofinformationarevitaltoinformationpropagation.News,aformofinformationwhichsatisfiesthelawofinformationpropagation,oughttobequalifiedfromabovetwoaspects.Theinherentvalueofnewsaredividedintothevalueofcontentandthequalityofpresentationwhichwehaveyzedabove.Wewilldiscusstheinformationflowefficiencyindetailasbelow.Aswehaveyzedabove,thetotalpropagationdelaytimeandthetotalpropagationratio,whichreflectstheflowefficiency,canbeobtainedthroughcomputercalculation.fytheefficiencyeff Wherei(i1,2)denotestheweight.1isnegativewhile2isInconclusion,thequalityofnewscanbecalculatedasquality1eff Wherei(i1,2)denotestheweightofithindicator.

denotesinformationefficiencyandcdenotestheinherentvaluewhichhas yzedForthedifferentsituation,wecansetdifferentthreshold.Wheninformationcanqualifiedasnews.

qualityTask2:PredicttheInformationCommunicationSituationforThemarketshareofdifferentmediaisanindicatortofytheinformationcommunication.ChangetheweightNinto1inequation(1)andleti0(i=,R,TV,I,M)tocalculatethepropagationratioN,whichisyzedinsection4.2.ComparedNwiththeactualmarketshareofnewsp r,wecometoyzetheerrorbetweenourmodelandthereality.Themarketshareforecastofothermediaisinthesameway.,wegiveoutthetotaltrendinmarketshareofsixmediaplotsasfollowsbasedregressionysis.Wechoosetworepresentativecountries:AmericaandChinaasstudyFigure10MediaMarketShareinAmerica Figure11MediaMarketShareinChinaThespecificvalueofmarketshareofdifferentmediaislistedinTable3andTable4.Table3MediaMarketSharein evision Table4MediaMarketShareinMediumNewspregraphRadio evisionInternet Wecanseethatthemarketshareoftraditionalmediasuchasnewsprandradioisdescendingallthetime.Onthecontrary,thenewmediasuchastheInternetandmobilephoneoccupymoremediashare.Therearesixmediaandsevenregions,forconcisewriting,wechoosethenewsprandtheInternetofAmericaforexample.Wecollectdatafromadiscussionprfromcommunic@tionsManagementInc[11]andBBCnews[12]. Figure12MarketshareofNewspr Figure13MarketshareoftheInternetFigure12showsthemarketshareofnewsprwhileFigure13showstheoneoftheInternet.Thebluefulllinerepresentstherealityandthegreendashedlinerepresentstheforecastbasedonthepastdata.Obviously,thegreenlineisclosetotheblueone,withtherelativeerrorisunder10%,showsthatourmodelisreliable.ThespecificerrorisshowninTable5.Table5TheErrorintheMarketShareofDifferentTrueAbsoluteRelativeTospecifytheinformationcommunicationsituationfortoday,wechooseaimportantworldeventaboutthedeathofBinLaden,whichhappensinMay1st,2011.Basedonourmodel,wecalculateinformationvalueasthenumberof8.3.Inordertoreducetheamountofcalculation,wechoosetocalculatethenumberofpropagationnodesinoneminutetoverifyourmodel.Followingareourresults.Figure14InformationSpreadWeobtainedthatthenumberofnodesis89.Accordingtothedatafrom[13],weknowthatthereare80peoplere-posttheinformationin ,whichissimilartoourevaluation.Fromaboveexample,wevalidateourmodel’sreliability.Task3:PredictthesituationaroundtheyearTherearesevenregions,thus,forconcisewriting,wechoosetworepresentativeregions,whereUnitedStatesandCongorespectivelyasourstudyobjects.Astimegoeson,oursocietytendstoinformation-overloaded,withmoreandmoreuselessinformationaroundus.Withthedevelopmentoftechnologyandmindschanging,itisobviouslythattheinherentvalueofinformationitselfdecreaseswhiletheefficiencyofinformationflowincreases.Toshowtherelationshipbetweenflowofspeedofinformation(eff)vsinherentvalueofinformation(c),whichhave fiedwedefinetheratiosass

Figure15Ratiochangesin Figure16RatiochangesinFigure15showstheratiochangesinAmerica.Thelinedecreasesquicklyintheearlytimeandreachesaplateauaroundtheyear2030,whichindicatesthepopularityrateoftheInternetwillbehighandstableatthatmoment.Theratiowillreachapproximaythenumberof14aroundtheyear2050.Inthesameway,Figure16showstrendofratiochangesinCongo,whichisdifferentfromtheoneofAmerica.Thelinedeclinesslowlyatandfallsfasterinabouttheyear2020,andfinallywillreachthenumberof30aroundtheyear2050,whichislowerthantheoneofAmerica.Thesefiguresindicatethedifferenceofsocialdevelopmentlevel.Accordingtoourconstructedmodel,thetotalpropagationtimeandthetotalpropagationradioarethetwoindicatorstoevaluatenetworks’capacities.Thedataareemployedtotrainthenetwork.Figure11and12showindicatorsandoftwocountriesfromtheyear1860totheyear2050below.Weassumethatthetativevalueofpropagationdelayisthenumberof1. Figure17CapacitieschangesinAmerica Figure18CapacitieschangesinCongoFigure17plotsthevariationtrendforthesetwoindicatorsofAmerica.Propagationdelaydecreaseswhilepropagationprobabilityincreasesonthecontrary.Therearetwotimepointinthegraph,the oneisaroundtheyear1870whenthenewsp rcometotheworld,whichimprovethespreadofinformationextremely.Thesecondoneistheyeararound2000becauseoftherapiddevelopmentofInternettechnology.Thesetwographsindicatethehistoryofmediaanditsimportanceinspreadinginformation.Figure18indicatescapacitiesinCongochangesinthesameTosumup,theglobalnetworkcapacityincreasesrapidlybeforetheyear2050andthedevelopedcountriesreachaplateauwhiledeveloco