| The development of the Internet and artificial intelligence technology has promoted changes in teaching methods.The arrival of the epidemic has accelerated the development of online teaching.Online live teaching has become a commonly used form of online teaching during the epidemic because of its strong interaction.A large amount of student data and course data are generated during the online live teaching,and valuable information can be extracted from it through machine learning technology,which can provide better support and help for the online live teaching.This thesis mainly constructs the curriculum profile of online live teaching from three aspects,including the division of online live teaching activities,student profile,and hierarchical curriculum knowledge graph.Presenting the overall situation of online live lessons through course profile can help teachers and evaluators to grasp the status of the course and conduct course evaluation,thereby helping teachers improve teaching and provide precise guidance to students.First of all,this thesis studies the multi-level division method of online live teaching videos.The online live class includes teaching activities such as course explanation,interactive Q&A,and in-class practice.Teaching activities consist of teaching behaviours,including teacher questions and student answers.The course content is composed of multiple knowledge points explained by the teacher.Traditional teaching analysis and evaluation requires manual viewing of teaching videos to record teaching behaviours and teaching activities,and then analyze the teaching arrangements to guide teachers to improve teaching plans.Traditional methods are time-consuming and labour-intensive,and cannot satisfy the situation of analyzing and evaluating large-scale courses.This thesis uses key frame and text recognition technology to automatically divide the time arrangement of knowledge points,uses speech recognition and text classification technology to automatically divide the teaching behavior and teaching activities in the live class,and carries out hierarchical integration of the division results.Then this thesis studies the construction of student profile tags and uses the learning behaviour data generated in the online live class to model the learning status and characteristics of students.This thesis constructs the student profile label from the student’s course participation,interaction level and the state of listening in the class.The clustering algorithm is used to divide the students’ degree of interaction and course participation.The listening status is analyzed through the students’ answer to the questions and the students’ behaviour in class.Students are divided into different groups according to the characteristics of students and assign profile labels.This could help the teacher to provide personalized guidance based on the characteristics of the students,and help students master the learning status and make improvements.Then this thesis studies the construction of hierarchical curriculum knowledge graph.The online teaching platform has abundant teaching resources,but the resources are scattered and the curriculum knowledge structure cannot be clearly presented.It is difficult for students to effectively use curriculum resources and master the connection of knowledge points.This thesis takes the database system course as an example to illustrate the method of constructing a hierarchical course knowledge graph through course teaching videos.Firstly,the knowledge points of the course are obtained through the named entity recognition algorithm,and then the relationship between the knowledge points is obtained through the relation extraction algorithm,and then the knowledge graphs of multiple courses are merged.Finally,this thesis describes the implementation plan of the curriculum profile system and displays the functions of the division of teaching activities,student profile and curriculum knowledge graphs.And the system has been verified through system testing and evaluation. |
PDF Full Text Request