Font Size: a A A

The Study Of Complex Network And Learning Behavior In MOOCs

Posted on:2021-01-03Degree:MasterType:Thesis
Country:ChinaCandidate:X W LiuFull Text:PDF
GTID:2370330647459958Subject:Computational Physics
Abstract/Summary:PDF Full Text Request
Relying on the booming development of the Internet,Massive Open Online Courses have realized knowledge transfer and educational innovation,and have expanded the boundary of higher education to a certain extent.However the low learning efficiency and high dropout rate in MOOCs are also widely criticized.Like most social systems,learners' online interactions can also be indicated as complex systems and network-related processes.Each learner represents individuals(nodes)in the network,and individual interactions constitute complex relationships in the network.At the same time,the prosperity of big data and artificial intelligencerelated technologies has also provided a variety of technical means for the study of MOOCs.Through physical modeling,analysis of individual interaction behaviors and dynamics of information transmission in the learning network can provide very valuable directions for the next development of MOOCs.This paper is the first to use social network analysis(SNA),information dissemination models and text classification techniques to study four different types of MOOCs courses in the Chinese community.We first introduce social network analysis,network vulnerability and information dissemination models study the differences in network structure and learning behavior of four different types of MOOCs courses.Then,based on simple machine learning technology,a basic text classifier was built to classify more than 100,000 posts into two categories related to the course content or not related to the course content.The parameters such as network module degree were introduced to compare the differences between the course network before and after text filtering and analyzes the roles played by these two types of posts in different types of courses.Results show that the network structure of different courses is determined by the learners' behavior which is closely related to the background of the learners,the characteristics of the course and the teacher's guidance;the roles of content and noncontent posts showing heterogeneity in different courses,non-content posts can promote global communication in science and engineering courses,but there is no such phenomenon observed in humanities courses.These results indicate that it is crucial to analyze learners' social behavior in different courses and establish different guiding mechanisms.At the end of this paper,we pointed out the possible directions for future research.
Keywords/Search Tags:Complex network, Simulations, Distance education and online learning, Data science applications in education
PDF Full Text Request
Related items