摘要
针对MOOC研究现状,结合MOOC平台多样、规模庞大、种类众多、学习者背景和目的各异等特点,以个性化预测研究为主要对象,分析和对比了作为研究基础的点击流日志、讨论区、课后任务、课程信息和学习者基本属性五类数据源的优劣,归纳了用于预测的基于活动、讨论区、社交和认知的模型。以《大学信息技术二(Python程序设计)》为例,构建和应用了个性化预测和干预模型。
According to the current situation of MOOC research,combining with the characteristics of MOOC platform of diversity,large scale,many kinds,and different backgrounds and purposes of learners’,this paper analyzes the advantages and disadvantages of five kinds of data sources,which as the research basis,namely click stream log,forum posts,assignments,course information and basic attributes of learners;and four models for personalized prediction including activity-based,discussion forum,social intercourse and cognitive are summarized.Taking"University Information Technology II(Python Programming)"as an example,the model of personalized prediction and intervention is constructed and applied.
作者
蒋翀
朱名勋
王点
Jiang Chong;Zhu Mingxun;Wang Dian(School of Information Science and Engineering,Hunan Woman's University,Changsha,Hunan 410004,China)
出处
《计算机时代》
2020年第8期10-13,共4页
Computer Era
基金
湖南省教育厅科学研究项目“基于移动微课的非结构化教育资源个性化推送算法研究”(16C0804)。
关键词
MOOC
个性化
预测模型
复制框架
MOOC
personalize
predictive model
replication framework