基于学习者行为特征的MOOCs学习伙伴推荐.pdfVIP

ISSN 1673-9418 CODEN JKYTA8 E-mail：fcst@vip．163．com JournalofFrontiersofComputerScienceandTechnology http：／／www．ceaj．org 1673--94182／015／09(01)—0·071·0·9 Tel：+86—10 doi：10．37788．issn．1673—9418．1405052 基于学习者行为特征的MOOCs学习伙伴推荐 徐 彬H，杨 丹 ，张 昱 ，李 封 ，高克宁 1．东北大学 计算中心，沈阳 110819 2．东北大学 信息科学与工程学院，沈阳 110819 Learners’ActivitiesBasedStudyBuddiesRecommendationTowardsM OOCs XU BinH，YANGDan2，ZHANGYu，LIFeng，GAOKening 1．ComputingCenter,Northeastern University,Shenyang110819，China 2．CollegeofInformationScienceandEngineering，NortheasternUniversiyt,Shenyang110819，China +Correspondingauthor：E—mail：xubin@mail．neu．edu．ca XU Bin，YANG Dan，ZHANG Yu，eta1．Learners’activitiesbased studybuddiesrecommendation towards MOOCs．JournalofFrontiersofComputerScienceandTechnology,2015，9(1)：71-79． Abstract：Therapiddeploymentofmassiveopenonlinecourses(MOOCs)hascreatedasurgeintheglobalconnec— tiviyt amongstudentsforeducationalpurposes．Withcontrasttolargescaleenrolledsutdents，almostallthecourses inMOOCshavealow completionrate．Oneoftheimportantreasonsofrthosewhodropoutthecour seislackof suitablesutdybuddiestocommunicateeachother．Thispaperanalyzessutdents’postedmessagesandtheirroles andbehaviorsinMOOCsdiscussionofrum，andbuildsstudents’behaviorfeaturemodelandrelationalnetworks betweensutdents．Toestimatesutdent’Srelevancywithrespectedtotopicwordsdictionarywhichisselectedaccording tocoursecontentinadvance，thispaperproposesanew topicmodelwiht fixedtopicdictionary,andfinallyrecom— mendssutdybuddiesofrsutdentby inferredtopicdisrtibution．ExperimentsoncrawledCourseradatasetdemon— stratethathtetopicmodelcanfindoutthosesutdentswithmoresimilariyt onsanletopicwords，andrecommend sutdybuddies．Andthiswillb