Research On The Construction Of Educational Knowledge Graph And Student Group Detection Method Based On MOOC Data

With the promotion of the "Internet + Education" model,the e-learning based on MOOC has developed rapidly,and massive amounts of educational big data with application value have emerged.In the pre-class stage of e-learning,teachers not only consider teaching objectives and strategies when preparing teaching plans,but also need to consider the individualized learning needs of students.However,the current preparation of teaching plans often relies on the teachers’ personal teaching experience and lacks a unified process.At the same time,the current MOOC model is also difficult to meet the individualized learning needs of students.As a kind of education big data,MOOC data can promote the intelligent transformation of the education system.For this purpose,the research contents of the pre-class stage in Hybrid MOOC mode are proposed: The construction method of multi-layer architecture knowledge graph based on MOOC data and the method of student groups detection based on MOOC data.(1)This paper proposes a method for constructing a multi-layer architecture knowledge graph based on MOOC data.It firstly defines the map of educational knowledge from the teaching resources and teaching subjects involved in teaching activities.Secondly,a multilayered educational knowledge graph TKS model is proposed,included the knowledge structure,construction process and construction method.Then it uses knowledge graph construction techniques such as entity extraction and relationship extraction to realize the knowledge graph in the TKS model and visualize it through the network analysis visualization tool Gephi based on the MOOC data.Finally,the teaching path is generated by combining the actual teaching scene and the constructed knowledge layer map.Teachers can not only clarify the knowledge points and their relationships when preparing the teaching plan,but also assist the teaching decision based on the teaching path.(2)This paper proposes a method of student group detection based on MOOC data.It firstly describes and formalizes the problem of student group detection based on MOOC data,then transforms it into a clustering problem.Secondly it describes the method of student group detection,including dual-view MOOC data acquisition and preprocessing process,similarity measurement,dual-view similarity network fusion clustering and evaluation of clustering result.Then the student groups are divided through numerical experiments.Teachers can rank the speed of learning knowledge points in each group after identifying the typical learning behaviors of them.Finally,the weights of knowledge points update and visual display by combining the learning behavior data and preference of each study group and the constructed knowledge layer map of the TKS model.Thus,teachers can not only understand the learning speed of knowledge points for each learning group,but also recognize the preference of knowledge points for them,so that they can carry out grouping and personalized teaching activities in class.This article provides the multi-layer structure knowledge graph construction method and student group detection method based on MOOC data.It can provide unified technical methods for teachers to design teaching plans in the pre-class stage and certain support for the construction of a new e-learning model based on MOOC model.