資源描述:
《畢業(yè)論文---試卷自動(dòng)生成系統(tǒng)的設(shè)計(jì)和實(shí)現(xiàn)》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在教育資源-天天文庫(kù)。
1、試卷自動(dòng)生成系統(tǒng)的設(shè)計(jì)和實(shí)現(xiàn)摘要隨著計(jì)算機(jī)技術(shù)的發(fā)展和智能優(yōu)化算法研究的深入,組卷系統(tǒng)的研究正被越來越多的專家學(xué)者所注意。它不僅涉及到組卷數(shù)學(xué)模型建立的問題,而且還包括相應(yīng)組卷算法的研究。組卷問題是一個(gè)在一定約束條件下的多目標(biāo)參數(shù)優(yōu)化問題,采用傳統(tǒng)的數(shù)學(xué)方法求解十分困難,自動(dòng)組卷的效率和質(zhì)量完全取決于試題庫(kù)的設(shè)計(jì)以及組卷算法的設(shè)計(jì)。因此如何設(shè)計(jì)一個(gè)算法從試題庫(kù)中既快又好地抽出一組最符合考試要求的試題,并基于此實(shí)現(xiàn)智能組卷系統(tǒng),是本文的研究目的。遺傳算法(GeneticAlgorithm)是一種模擬大自然生物進(jìn)化過程的智能算法,它以其簡(jiǎn)單、魯棒性強(qiáng)、全局
2、尋找以及不受搜索空間限制性條件約束等特點(diǎn)而備受關(guān)注。遺傳算法的群體搜索策略為多目標(biāo)優(yōu)化提供了非常合適的解決方案,因此將其應(yīng)用于組卷問題能取得良好的效果。本文為實(shí)現(xiàn)試卷自動(dòng)生成系統(tǒng),從教育測(cè)量學(xué)、人工智能、軟件工程等領(lǐng)域出發(fā),基于改進(jìn)遺傳算實(shí)現(xiàn)了自動(dòng)組卷算法,建立了試卷自動(dòng)生成系統(tǒng),驗(yàn)證了該算法的可行性和有效性。關(guān)鍵詞:遺傳算法,自動(dòng)組卷,數(shù)學(xué)模型,目標(biāo)約束2DESIGNANDIMPLEMEMTATIONOFTESTPAPERAUTOMATICGENERATINGSYSTEMABSTRACTWiththedevelopmentofcomputertech
3、nologyandintelligentin-depthstudyofoptimizationalgorithm,theresearchofautomaticgeneratingtestpapersystemisbeingpaidattentiontobymoreandmoreexpertsandscholars.Itnotonlyreferstotheestablishmentoftestpapergeneratingmathematicalmodel,butalsoincludestheresearchofcorrespondingalgorith
4、ms.Thetestpaperauto-generatingisanoptimizedproblemtomulti-objectiveparameterwithacertainconstraints.Theoptimizationisimplementedverydifficultybytraditionalmathematicalmethods.Theefficiencyandqualityofautomaticgeneratingtestpaperisabsolutelydeterminedbythedesignoftestquestionsdat
5、abaseandcorrespondingautomaticgeneratingtestpaperalgorithm.Sohowtodesignaalgorithmtoselectagroupoftestquestionsmostmatchingtherequirementsofexaminationeffectivelyandefficiently,whilebasingonittorealizetheintelligenttestpapergeneratingsystem,whichisthepurposeofthispaper.Previousp
6、apergeneratingalgorithmsaremostlybasedonrandomselectivestrategyandrecall-teststrategy.Theformeriseasytorealizewith2hightimecomplexity,thelatteroccupieshighspacecomplexitywhiletotalquestionnumberiscomparativelylarge,neitherhastheintelligence.GeneticAlgorithm(GA)isanintelligentalg
7、orithm,whichsimulatesthenaturalprocessofbiologicalevolution.Itisbeingpaidmoreandmoreattentiontowiththecharacteristicsofsimple,strongrobustness,globalsearchandunfetteredbytherestrictiveconditionsofsearchspace.ThepopulationsearchstrategyinGAprovidesaverysuitablesolutionformulti-ob
8、jectiveoptimization,soapplyingittotheissueoftes