【摘要】 基于大数据驱动思想,笔者利用复杂网络理论及其社团特性,提出一种SPOC教学中学习小组及学习领袖的自动发现机制,并且可实现学习小组及小组内的成员的数量调整的动态自适应。首先,通过融合学习者的成绩和性格特质的复合欧式距离构建用户模型,进而构建学习者网络;其次,利用Fast community算法实现网络的社团划分;最后,以1个社团作为1个学习小组,并将其中介数最大的节点作为该组的学习领袖。与单独学习相比,实施所提方法后,学习者的平均成绩提高4.7%。更多还原

【Abstract】 Based on Big data-drive,using complex network theory and community,this paper proposes a discovering scheme of learning groups and its leaders for SPOC,which can adaptive adjustment the number of learning group and its member:Firstly,constructing a user model with personality based on the Euclidean distance,which integrated personality and academic record,between learners,and then constructing the learner network.Secondly,we divide some communities among this network by Fast community.Thirdly,one community is referred as a learning team,and the node with maximal betweeness among it is selected as group’s leaders.Through analyzing the SPOC data,we find that the community does exist in the learner network.And the discovering of the community with single node can help teachers find the learners who need special attention in teaching.Compared to the individual learning,the implementation of the proposed method can improve the average score with 4.7%.更多还原