A Research Based On MOOC Course Comment For Learning Behavior Analysis

Since 2012,MOOC has mushroomed in the field of online education in China.The massive,open and online superiority has received widespread attention and has been constantly changing and optimizing offline education models.As an indispensable part of the MOOC learning system,student comment information in the MOOC classroom can effectively present the student's learning results and observe the student's learning status.It is also easy to reflect the teaching rules,and enhance interaction between teachers and students,in order to create a better teaching and learning experience.In the environment of a computer,massive course comment is easy to store and gather.Meanwhile,combined with statistics and text processing technology,the quantized course comment information can be used to calculate and analyze.Considering that existing literatures only performs statistical data on quantitative learning comment information and cannot deeply extract textual information from the content.So,this article integrates the textual processing of course comment content on the basis of course comment statistics,and experiments prove that the innovation is effective for further study of MOOC learning behavior analysis.Based on this,this paper selected 34 courses from the Chinese University MOOC platform and captured 510,000 classroom question-student comment data in the class discussion area.From the two levels of curriculum and students,we will discuss the MOOC learning behavior rules including explicit behavior and implicit behavior.On the one hand,The comment time,the number of participants,the total amount of comment and the total number of subjects were used as the observation variables.Two levels of statistical indicators of participation quality and activity are designed for course comment records.Statistical methods were used to calculate the student participation rate,time distribution,student proportion distribution,student certificate qualification rate,and student learning time allocation rate comment in class to track the participation rate and activity change of the course learning behavior,and the quality of participation and active rate of student learning behavior.On the other hand,in response to the course comment content data,combined with text similarity calculation technology,keyword extraction technology and sentiment analysis technology in the field of natural language processing,the course comment texts of the two parts of the curriculum and students are deeply excavated and discoursed analyzed,in order to observe the course comment behaviors and students' learning behavior law in classroom discussion areas,and to observe the learners' learning behavior from the perspective of different disciplines.First,from the curriculum level.Starting from the two dimensions of the student comment record and comment content,this paper calculated the student participation rate,comment participation rate,time distribution and student proportion distribution,etc.,in order to track the change of student participation and activity.Text similarity calculation,keyword extraction and sentiment analysis,three types of text analysis techniques are used to observe the discourse distribution of course comment at the curriculum level,and to observe MOOC course learning behavior differences between different areas.The experiment found that the curriculum comment showed a trend of"big first and then small" in the distribution of time.The degree of participation and activity of different discipline types were uneven,and the overall level was low.But the content of comment and the topic are closely related,and most course comments perform positive emotions.Second,from the student level.From the learner's personal homepage,the comment of learning comment and comment are taken into account.Through the calculation of the qualification rate of the certificate,the rate of comment approval,the proportion of students' learning time,etc.,the learner's quality of class learning and the degree of activity are observed,and the related natural language processing technology is used to focus on the topic distribution of students.Experiment found that the total comment of students was not coordinated with the proportion of the number of registered courses,and the comment activity and learning achievement rate are obviously not high,and the degree of interaction between students is also low.However,the student's comment text is mainly positive,and show unique styles in terms of curriculum and themes.Finally,we have implemented a learning behavior analysis system for MOOC course comment.The system can obtain course comment data and students' learning information,and automatically calculate course comment time distribution,student distribution,sentiment distribution,and keyword distribution for each topic in the course,as well as outline the user's portraits and automatically calculate student learning information,comment emotion distribution and keyword distribution,all-round to clearly present the MOOC courses and students' learning behaviors.