Analyzing And Quantifying The Attraction Of Online Course

Along with information technology more widely applied in education,as a new form of education in the information age,massive open online courses have penetrated into our life.Online learning is not limited by time and region,so as to enhance interactive experience and realize personalized learning of online learning.But learners are prone to problems such as insufficient investment of time,the lack of planning,assault browsing,random termination,high drop-out rates and low pass rate.The quality analysis method of the traditional classroom analysis method(experts,based on the standard,periodic)is difficult to adapt to analyze the attraction of online course videos.Need to come up with new ways,new method to study online course video attraction.This study deepens the deep and multi-dimensional understanding of the law of online learning behavior,which is more conducive to improving the quality of online courses.The online learning platform has accumulated massive learning behavior data in the process of development.Mining the value of online learning behavior data and exploring the rules of online learning behavior are effective ways and methods to solve the attraction of online courses videos.The study of the attraction of online course video is of great significance in sociology and education.Therefore,According to the data characteristics of online courses,this paper adopts data-driven research method to complete the following work:1.Analyzing the attraction of online course.Through statistical analysis,a single video attraction factor R is established,this study is based on multiple regression algorithm to study the mathematical mechanism of single video attraction affected by video attributes.It is found that the attractiveness of a single video is negatively linearly correlated with video duration and video release time.When comparing the absolute value of correlation coefficient,video duration has a stronger linear correlation with R than video release time.On this basis,the interval of optimal video time length was estimated by statistics of the most popular video,and the reliability was verified by Bootstrap method.Finally,the time scaling factor of video is given to estimate the optimal time of video changing with the release time.These results can help video producers improve the design of video and make their courses more attractive.2.Quantifying the attraction of online course.This system is presented by a bipartite network after we identify whole-willinglearners,nodes being the videos and the learners.In this network,each learner node has a L-value which reflects its learning effectiveness and each video node has a V-value which reflects the contribution of the video to the attraction of the course,there is a mutual strengthening relationship between the two sets.In order to quantify the attraction of online course,we propose a bipartite network mutual reinforcement algorithm.The method grades the learning effectiveness of each learner.Interestingly,these scores can be devided into two sets,and each set obeys normal distribution.Thus,we propose a method to recognize deep-learners in the whole-willing-learners.Finally,we define the index P to quantify the attraction of online course,and obtain that the project-driven learning curriculum is more attractive to learners.The bipartite graph mutual reinforcement algorithm is suitable for the evaluation of node importance in a bipartite network.