搜索
英文論文寫作與投稿經驗--黃國禎1英文論文寫作與投稿經驗國立臺南大學數位學習科技系黃國禎英文論文寫作與投稿經驗--黃國禎2一般期刊之評審要點學術價值(創新性)應用價值學理根據與觀點之正確性文章組織結構研究方法之嚴謹性題目合宜文章長度恰當格式正確用詞的正確性及文章的流暢度英文論文寫作與投稿經驗--黃國禎3研究的觀念、態度與方法觀念研究是有趣的態度充滿好奇心、全力以卦方法功勞＋苦勞＝創新＋實証英文論文寫作與投稿經驗--黃國禎4研究主題的選擇儘量針對同一主題以不同的方法深入探討Gwo-JenHwang,Tsung-LiangLin,BertrandM.T.Lin(2004),“AnEffectiveApproachforTest-SheetCompositionfromLarge-ScaleItemBanks”,acceptedbyComputers&Education.(SSCI)Gwo-JenHwang(2003),“ATestSheetGeneratingAlgorithmforMultipleAssessmentRequirements”,IEEETransactionsonEducation,Vol.46,No.3,pp.329-337.(SCIandEI)Gwo-JenHwangetal.(2005),“OntheDevelopmentofaComputer-AssistedTestingSystemwithGeneticTestSheet-GeneratingApproach”,acceptedbyIEEETransactionsonSystems,Man,andCybernetics:PartC(SCI,EI).Gwo-JenHwang,Peng-YengYinandShu-HengYeh(2005),“ATabuSearchApproachtoGeneratingTestSheetsforMultipleAssessmentCriteria”,acceptedbyIEEETransactionsonEducation.(SCI,EI)英文論文寫作與投稿經驗--黃國禎5論文架構-以系統為主的研究標題及作者:Titleandauthors摘要:Abstract簡介:Introduction相關研究:RelevantResearch(LiteratureReview)系統架構:SystemStructure研究方法(演算法):XXXApproach(Algorithm)系統製作、實驗及評量:ExperimentsandEvaluation結論參考文獻英文論文寫作與投稿經驗--黃國禎6論文架構-以方法為主的研究標題及作者:Titleandauthors摘要:Abstract簡介:Introduction相關研究:RelevantResearch(LiteratureReview)問題描述:ProblemDefinition研究方法(演算法):XXXApproach(Algorithm)系統製作、實驗及評量:ExperimentsandEvaluation結論參考文獻英文論文寫作與投稿經驗--黃國禎7論文寫作的建議順序相關研究問題描述或系統架構研究方法系統製作、實驗及評量簡介摘要結論決定題目英文論文寫作與投稿經驗--黃國禎8論文題目10-15字能立即呈現研究的目的及貢獻DevelopmentofaTestingSystem(X)ANewTestSheetGeneratingMethod(X)ANovelApproachtoComposingTestsheetsforMultipleAssessmentCriteriainBuildingTestingSystems(O)DevelopmentofATestingSystemtomeetMultipleAssessmentRequirements(O)英文論文寫作與投稿經驗--黃國禎9相關研究避免嘗試改寫中文版的相關研究，直接重寫會比較快。先找好10-20篇最近十年相關的文獻。挑選2-3篇最直接相關的文獻，參考其literaturereview的內容，來描述問題的形成動機。再參考其他文獻的Abstract描述，依年代分段敘述最近十年的發展狀況，約1000-1500字重點：Tellastory(加一些說明將這些內容連貫起來）英文論文寫作與投稿經驗--黃國禎10Inrecentyears,researchershavedevelopedvariouscomputer-assistedtestingsystemstoevaluatestudents’learningstatusmoreprecisely.Forexample,FeldmanandJonesattemptedtoperformsemi-automatictestingofstudentsoftwareunderUnixsystems[5],Rasmussenetal.exercisedasystemtoevaluatestudentslearningstatusoncomputernetworksbytakingtheirprogressintoconsideration[22];furthermore,ChouproposedtheCATESsystem[3],whichisaninteractivetestingsystemdevelopedinacollectiveandcollaborativeprojectwiththeoreticalandpracticalresearchoncomplextechnology-dependentlearningenvironments.Unfortunately,althoughmanycomputer-assistedtestingsystemshavebeenproposed,fewofthemhaveaddressedtheproblemoffindingasystematicapproachforcomposingtestsheetssatisfyingmultipleassessmentrequirements.Mostoftheexistingsystemsconstructatestsheetbymanuallyorrandomlyselectingtestitemsfromtheiritembanks.Suchmanualorrandomtestitemselectionstrategiesareinefficientandarehardlyabletomeetmultipleassessmentrequirementssimultaneously.英文論文寫作與投稿經驗--黃國禎11Somepreviousinvestigationsshowedthatawell-constructedtestsheetnotonlyhelpsevaluatethelearningstatusofthestudents,butalsofacilitatesthediagnosisoftheproblemsembeddedinthelearningprocess[13,14,17].Thatis,itisverycriticaltoselectpropertestitemstoconstituteatestsheetthatmeetsmultipleassessmentcriteria,includingtheexpectedtimeneededforansweringthetestsheet,thenumberoftestitems,thespecifieddistributionofcourseconceptstobelearned,andthemostimportant,themaximizationoftheaveragedegreeofdiscrimination[20].Sinceitisdifficulttosatisfymultiplerequirements(orconstraints)inselectingtestitems,mostcomputerizedtestingsystemsgeneratetestsheetsinarandomfashion(whichwillbecalled“randomselection”throughoutthispaper)[16].In[15],amultiplecriteriatestsheetgeneratingproblemisformulatedasadynamicprogrammingmodel[11]tominimizethedistancebetweentheparameters(e.g.,discrimination,difficulty,etc.)ofthegeneratedtestsheetsandtheobjectivevaluessubjecttothedistributionofconceptweights.英文論文寫作與投稿經驗--黃國禎12Althoughthedynamicprogrammingapproachhastakenmultiplerequirementsintoconsideration,inpracticalapplications,morecriterianeedtobeaddressed.Forexample,ateachermightliketoassignarangeoftesttimesinsteadofgivingafixedtesttimeandusuallyateacherwillassignanexpectedlowerboundforeachconceptweightinsteadofgivingadistributionofconceptweights.Moreover,thegoalofagrouptestistodiscriminatethestatusofthestudents.Thisimpliesthatthediscriminationdegreeoftheentiretestsheetneedstobemaximizedasmuchaspossible.Anothercriticalissuearisingfromtheuseofdynamicprogrammingapproachisprobablytheexceedinglylongexecutiontimerequiredforproducingoptimalsolutions.Asthetime-complexityofthedynamicprogrammingalgorithmisexponentialintermsofinputdata,theexecutiontimewillbecomeunacceptablylongifthenumberofcandidatetestitemsislarge.英文論文寫作與投稿經驗--黃國禎13Tocopewiththeseproblems,researchersattemptedtoformulateanewtestsheet-generatingproblembyoptimizingthediscriminationdegreeofthegeneratedtestsheetswithaspecifiedrangeofassessmenttimeandsomeothermultipleconstraints[16].However,thenewformulationisevenmorecomplex,previousinvestigationsshowedthatitisdifficulttoefficientlyfindnea