Abstract — An attempt is made to apply Ant Colony Optimization (ACO) heuristics to an E-learning problem: the pedagogic material of an online teaching website for high school students is modelled as a navigation graph where nodes are exercises or lessons
VIII.CONCLUSIONSANDPERSPECTIVE
Paraschoolwantedasmartautomaticsystemthatcouldadapttodifferentuserswithoutmanualintervention,whichwouldbetotallyunrealistictoenvisageon10000students.Theant-basedsystemdescribedinthispapernotonlyofferssuchautomaticfeaturesbygraduallymodifyingpedagogicpathssuggestedbyteachersusingcollectiveexperienceandbymakingthestructureindividual-speci cthankstovariablessuchasHbutalsocomesupwithemergentinformationsthatcanbeusedasare nedauditingtooltohelpthepedagogicalteamidentifythestrengthsandweaknessesofthesoftwareandpedagogicmaterial.
[16]V.Ramos,J.J.Merelo,Self-OrganizedStigmergicDocumentMaps:
EnvironmentasaMechanismforContextLearning.In1stSpanishConferenceonEvolutionaryandBio-InspiredAlgorithms,E.Alba,F.Herrera,J.J.Mereloetal.(Eds.),pp.284-293,Spain,6-8Feb.2002.[17]M.Resnick,Turtles,TermitesandTraf cJamsExplorationsinMas-sivelyParallelMicroworlds.ComplexAdaptiveSystemsseriesMITPress,1994
[18]D.Smith,G.Hardaker.E-LearningInnovationthrough
theImplementationofanInternetSupportedLearningEnvironment.InEducationalTechnology&Society3(3)2000.ifets.ieee.org/periodical/vol32000/e04.pdf[19]T.St¨utzleandM.Dorigo.ACOAlgorithmsfortheTravelingSalesman
Problem.InProceedingofEUROGENconferenceMMakela,KMietti-nen,PNeittaanmaki,JPeriaux(Eds),JohnWiley&Sons,1999,ISBN:0471999024.[20]T.St¨utzleandH.Hoos.TheMax-Minantsystemandlocalsearchforthe
Fromamoretheoreticalstandpoint,thisworkcanbeseenasanewtakeonInteractiveEvolutionaryComputationwherethesolutiontoaproblemisgraduallyconstructedandmodi edbymultipleinteractingentitieswithdifferentandpossiblyoppo-sitegoals.Acreativeandrobustcompromisecanbereachedthatbalancesallthein uencesandconstraints,whichallowsallparticipatingentitiestobene tfromanemergentcultureandtoenhancetheirdecisionmakingprocessesaccordingly.Thissuggestagreatdealofnewandexcitingapplicationsinthe eldofCollectiveCognitionModellingandCollectiveEvolutionaryDesign.
REFERENCES
[1]E.Bonabeau,M.Dorigo,andG.Theraulaz,SwarmIntelligence:From
NaturaltoArti cialSystems,AvolumeintheSantaFeinstituteintheSciencesofComplexityOxfordUniversityPress1999.ISBN0-19-513159-2.
[2]J.Bruner.ConstructivistTheory.doc.guandang.net[3]A.Colorni,M.Dorigo,V.Maniezzo.Distributedoptimizationbyant
colonies.InProceedingsoftheFirstEuropeanConferenceonArti cialLife,pages134-142.Elsevier,1992.
[4]J.L.Denebourg.Theself-organizingexploratorypatternofthearhentine
ant.JournalofInsectBehavior,3:159-168,1990.
[5]M.Dorigo.Optimization,LearningandNaturalAlgorithms.PhDThesis,
DipartimentodiElectronica,PolitecnicodiMilano,IT,1992.
[6]M.Dorigo,G.DiCarroandL.M.Gambardella.AntAlgorithmsfor
DiscreteOptimization.Ariti cialLife,Vol.5,No.3,pp.137-172,1999.[7]D.E.Goldberg,TheRace,theHurdle,andtheSweetSpot:Lessons
fromGeneticAlgorithmsfortheAutomationofDesignInnovationandCreativity,ILLIGALReportNo98007,UrbanaIL.
[8]D.A.Goldberg.GeneticAlgorithmsinSearch,Optimization,andMa-chineLearning.Addison-Wesley,January1989.
[9]D.E.Goldberg,K.Deb,Acomparativeanalysisofselectionschemes
usedingeneticalgorithms,inFOGA91,vol1,pp69-93,1991,alsoTCGAReport90007.
[10]W.Jansen.TheAddedValueofE-learning.InformingSci-ence.InSITE-”WhereParallelsIntersect”.June2002.ecom-merce.lebow.drexel.edu/eli/2002Proceedings/papers/Janse124Added.pdf[11]D.L¨age,S.Daub.TheResurrectionofIntelligentTutorialSystems:
ANewWayofFactualKnowledgeDiagnosis.InternationalJournalofPsychology,35issue3/4,136,2000.
[12]D.MerkleandM.Middendorf,ModellingtheDynamicsofAntColony
OptimizationAlgorithms.InProceedingsoftheGeneticandEvolution-aryComputationConference(GECCO-2002),NewYork,2002.
[13]M.Middendorf,F.Reischle,andH.Schmeck.MultiColonyAntAlgo-rithms.JournalofHeuristics(specialissueonParallelMetaheuristics),8(3):305-320,2002.
[14]SeymourPapert.“PapertonPiaget.”InTimemagazine’sspecialissue
on”TheCentury’sGreatestMinds,”p.105,March29,1999.
[15]V.Ramos,F.Almeida.Arti cialAntColonesinDigitalImageHabitants
-AMassBehaviourEffectStudyonPatternRecognition.InProc.ofSIARP’2000-5thIberoAmericanSymposiumonPatternRecognition,pp.69-78,Lisbon,Portugal,11-13Sep.2000.
travelingsalesmanproblem.InT.Baeck,Z.MickalewiczandX.Yao,editors,ProceedingsofIEEE-ICEC-EPS’97InternationalConferenceonEvolutionaryComputationandEvolutionaryProgrammingConference,pages309-314.IEEEPress,1997.
[21]T.Teich,M.Fischer,A.Vogel,J.Fischer.AnewAntColonyAlgorithm
fortheJobShopSchedulingProblem.InProceedingsoftheGeneticandEvolutionaryComputationConference,SanFrancisco,California,July7-11,2001,p.803
[22]T.L.Wentling,C.Waight,J.Gallaher,J.LaFleur,C.Wang,A.
Kanfer,E-learning-AReviewofLiterature.KnowledgeandLearningSystemsGroup,UniversityofIllinoisatUrbana-Champaignlearn-ing.ncsa.uiuc.edu/papers/elearnlit.pdf
[23]WhiteT.,PagurekBandOppacherF.,ASGA:ImprovingtheAnt
SystembyIntegrationwithGeneticAlgorithms.InProceedingsoftheSymposiumonGeneticAlgorithms(SGA’98),July,1998.
[24]D.Whitley.TheGENITORalgorithmandselectionpressure:Whyrank–
basedallocationofreproductivetrialsisbest.InSchaffer(1989),pages116–121
热门文档