Font Size: a A A

Modeling And Analyzing MOOC Online Learning Behavior

Posted on:2019-03-26Degree:MasterType:Thesis
Country:ChinaCandidate:G C ZhouFull Text:PDF
GTID:2370330611493147Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
Massive Open Online Courses(MOOC)is a new pattern of education that combined technology and education,and it's an important part of “Internet + education”,too.MOOC is obviously different from traditional teaching pattern,especially in its informational characteristics,which can accumulate amounts of information of learners.Through analyzing these information,it can not only help to understand online learning behaviors and analyze its mechanism,but also help to explore a more efficient way for dispersing knowledge,hoping to provide theoretical support for online education's development and traditional education's reformation.We used the relevant knowledge of statistics and information theory to carry out the following research work:(1)Analyzing learner's characteristics and classifying the behavior pattern.Based on the attribute data,this paper analyzed the learner's attribute characteristics and give pictures to the learners.It found that different areas have different MOOC learners,and the difference is homogenous among different courses.This paper analyzed learner's behavioral characteristics from the behavioral data,and verified the characteristics of “high dropout rate and low participation”which is very common in MOOC.It also found that the average viewing degree of the course was negatively correlated with the average duration.Furthermore,this paper used the parameter and non-parametric methods to classify learner behavior patterns,and then explored the relationship between learner behavior characteristics and learning effects.(2)Measurement of video attraction.Based on learner's behavior data,taking video playback volume and learner's harvest into consideration and combined the H-index evaluation theory?Information entropy theory,this paper set up a mixed quantitative index to measure video attraction,and gave an scientific explanation of the index by analyzing learner's viewing behavior characteristics.(3)This paper analyzed the cumulative viewing time of online courses based on Survival analysis,it found that function which describes risk of the cumulative viewing time is controlled by Bathtub curve,and its tail has Lindy effect.This paper established two differential equations to describe the growth mode of the cumulative viewing time.Experimental result shows that the solution could recur the characteristics of the density function of the cumulative viewing time and the characteristics of the corresponding risk function.On one hand,this model explains mechanism of online learning behaviors,on the other hand,it provides a new method to examine the effect of teaching improvement.
Keywords/Search Tags:Massive Open Online Courses, Behavioral patterns of learning, Video attraction, Survival analysis
PDF Full Text Request
Related items